PRX QUANTUM 4, 027001 (2023)

Roadmap

Quantum Simulation for High-Energy Physics

Christian W. Bauer,1,* Zohreh Davoudi ,2,† A. Baha Balantekin,3 Tanmoy Bhattacharya,4
Marcela Carena,5,6,7,8 Wibe A. de Jong,1 Patrick Draper,9 Aida El-Khadra,9 Nate Gemelke,10
Masanori Hanada,11 Dmitri Kharzeev,12,13 Henry Lamm,5 Ying-Ying Li,14,15 Junyu Liu ,16,17
Mikhail Lukin,18 Yannick Meurice,19 Christopher Monroe,20,21,22,23 Benjamin Nachman,1
Guido Pagano,24 John Preskill,25 Enrico Rinaldi,26,27,28 Alessandro Roggero,29,30 David I. Santiago,31,32
Martin J. Savage,33 Irfan Siddiqi,31,32,34 George Siopsis,35 David Van Zanten,5 Nathan Wiebe,36,37
Yukari Yamauchi,2 Kübra Yeter-Aydeniz,38 and Silvia Zorzetti6
1
Physics Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, USA
2
Department of Physics, Maryland Center for Fundamental Physics, and the NSF Institute for Robust Quantum
Simulation, University of Maryland, College Park, Maryland 20742, USA
3
Department of Physics, University of Wisconsin, Madison, Wisconsin 53706, USA
4
T-2, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA
5
Fermi National Accelerator Laboratory, Batavia, Illinois 60510, USA
6
Enrico Fermi Institute, University of Chicago, Chicago, Illinois 60637, USA
7
Kavli Institute for Cosmological Physics, University of Chicago, Chicago, Illinois 60637, USA
8
Department of Physics, University of Chicago, Chicago, Illinois 60637, USA
9
Department of Physics and Illinois Center for the Advanced Study of the Universe and Illinois Quantum
Information Science and Technology Center, University of Illinois, Urbana, Illinois 61801, USA
10
QuEra Computing Inc, Boston, Massachusetts 02135, USA
11
Department of Mathematics, University of Surrey, Guildford, Surrey GU2 7XH, United Kingdom
12
Center for Nuclear Theory, Department of Physics and Astronomy, Stony Brook University, New York
11794-3800, USA
13
Department of Physics, Brookhaven National Laboratory, Upton, New York 11973, USA
14
Peng Huanwu Center for Fundamental Theory, Hefei, Anhui 230026, China
15
Interdisciplinary Center for Theoretical Study, University of Science and Technology of China, Hefei, Anhui
230026, China
16
Pritzker School of Molecular Engineering, Chicago Quantum Exchange, and Kadanoff Center for Theoretical
Physics, University of Chicago, Illinois 60637, USA
17
qBraid Co., Harper Court 5235, Chicago, Illinois 60615, USA
18
Department of Physics, Harvard University, Cambridge, Massachusetts 02138, USA
19
Department of Physics and Astronomy, University of Iowa, Iowa City, Iowa 52242, USA
20
Duke Quantum Center, Duke University, Durham, North Carolina 27701, USA
21
Department of Electrical and Computer Engineering, Duke University, Durham, North Carolina 27708, USA
22
Department of Physics, Duke University, Durham, North Carolina 27708
23
IonQ, Inc., College Park, Maryland 20740, USA
24
Physics and Astronomy Department, Rice University, Houston, Texas 77005, USA
25
Institute for Quantum Information and Matter, California Institute of Technology, Pasadena, California 91125,
USA
26
Physics Department, University of Michigan, Ann Arbor, Michigan 48109, USA
27
Theoretical Quantum Physics Laboratory, Center for Pioneering Research, RIKEN, Wako, Saitama 351-0198,
Japan
28
Interdisciplinary Theoretical and Mathematical Sciences Program (iTHEMS), RIKEN, Wako, Saitama
351-0198, Japan
29
Dipartimento di Fisica, University of Trento, via Sommarive 14, Povo, Trento I–38123, Italy

*
Corresponding author. cwbauer@[Link]
†
Corresponding author. davoudi@[Link]
Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further
distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.

2691-3399/23/4(2)/027001(70) 027001-1 Published by the American Physical Society

CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

30
INFN-TIFPA Trento Institute of Fundamental Physics and Applications, Trento, Italy
31
Quantum Nanoelectronics Laboratory, Department of Physics, University of California at Berkeley, Berkeley,
California 94720, USA
32
Computational Research Division, Lawrence Berkeley National Lab, Berkeley, California 94720, USA
33
InQubator for Quantum Simulation (IQuS), Department of Physics, University of Washington, Seattle,
Washington 98195, USA
34
Materials Sciences Division, Lawrence Berkeley National Lab, Berkeley, California 94720, USA
35
Department of Physics and Astronomy, University of Tennessee, Knoxville, Tennessee 37996-1200, USA
36
Department of Computer Science, University of Toronto, Toronto, Ontario M5S 2E4, Canada
37
Pacific Northwest National Laboratory, Richland, Washington 99354, USA
38
Emerging Technologies and Physical Sciences Department, The MITRE Corporation, 7515 Colshire Drive,
McLean, Virginia 22102-7539, USA

(Received 26 July 2022; revised 18 January 2023; accepted 6 March 2023; published 3 May 2023)

It is for the first time that quantum simulation for high-energy physics (HEP) is studied in the U.S.
decadal particle-physics community planning, and in fact until recently, this was not considered a main-
stream topic in the community. This fact speaks of a remarkable rate of growth of this subfield over the past
few years, stimulated by the impressive advancements in quantum information sciences (QIS) and associ-
ated technologies over the past decade, and the significant investment in this area by the government and
private sectors in the U.S. and other countries. High-energy physicists have quickly identified problems
of importance to our understanding of nature at the most fundamental level, from tiniest distances to cos-
mological extents, that are intractable with classical computers but may benefit from quantum advantage.
They have initiated, and continue to carry out, a vigorous program in theory, algorithm, and hardware co-
design for simulations of relevance to the HEP mission. This Roadmap is an attempt to bring this exciting
and yet challenging area of research to the spotlight, and to elaborate on what the promises, requirements,
challenges, and potential solutions are over the next decade and beyond.

DOI: 10.1103/PRXQuantum.4.027001

I. INTRODUCTION exponentially more compactly on a quantum simulator,

Quantum simulation refers to simulating a complex but also quantum operations retain quantum correlations
quantum system using another quantum system that can be and effectively perform massively parallel computations
fabricated in a laboratory, and be sufficiently isolated and on the encoded wave function. Harnessing such a signif-
controlled to maintain quantum coherence. The simulator icant computing resource amounts to finding a mapping
can closely resemble the degrees of freedom of the tar- between the dynamics of the target system and those that
get system, and its dynamics can be engineered to closely can be encoded in the simulator via quantum algorithms.
follow those of the simulated theory, in which case it is One desires such algorithms to require resources that scale
said to operate in an analog mode. The simulator can also subexponentially with the system size and the accuracy
be developed in such a way that the simulation will be desired. Furthermore, efficient error-correction algorithms
agnostic to the underlying physical architecture, and is in the far term, and effective error-mitigation techniques
engineered to implement a set of universal and elemen- in the near term, are needed to overcome the effect of
tary operations on an array of few-state quantum units, quantum decoherence in realistic implementations. Finally,
prototypically two-state “qubits,” in which case it is said an integrated hardware-software-compiler stack is desired
to operate as a digital quantum computer. It can also com- such that the users with applications can easily interact
bine features of the two modes for more adaptability and with the simulator, without significant barriers in trans-
efficiency. The simulated theory often describes a quan- lating between the simulated theory’s language and the
tum many-body system that is hard to simulate classically machine’s language.
as the Hilbert spaces are exponentially large in particle The development and advancement of the quantum-
number or system size, prohibiting exact or approximate simulation program in HEP requires all the above topics
classical Hamiltonian-simulation methods, or other meth- to be considered in the context of the target problems in
ods that are stochastic in nature and may suffer from sign HEP. Perhaps more importantly in the short term is to
problems. Not only can the Hilbert spaces be encoded identify such target problems, i.e., those that are likely

027001-2
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

Analog Simulators

Digital Computers
Collider
NISQ-Era Simulations Phenomenology

Software and Compiler Matter in and out

of Equilibrium
Quantum Simulators Neutrino
(Astro)physics
Quantum Simulation for
High-Energy Physics Early Universe
and Cosmology
Quantum Ecosystem
Quantum Gravity

Co-design and
Accessibility Physics Drives
Workforce
Development Underlying Simulations

Strategic
Partnerships
Quantum Field Theory Effective Field Theory
Simulations Simulations

FIG. 1. Aspects of a quantum simulation program for HEP that are studied in this Roadmap.

to remain intractable with current theoretical and com- Given such a multipronged and multidisciplinary area of
putational methods, but have the potential to be solved research and development in quantum simulation of HEP,
with quantum simulators. Once such problems are identi- a new and skilled generation of workforce will need to
fied, the underlying theoretical framework must be adopted be trained and empowered, and strong partnerships among
properly and the corresponding quantum simulation algo- universities, national laboratories, and technology compa-
rithms need to be developed. Since the underlying theoret- nies to be pursued. This will allow theoretical, algorithmic,
ical framework in HEP is generally quantum field theories and experimental lines of this research to be advanced
(QFTs), the aim becomes quantum simulating quantum quickly and simultaneously.
fields and their interactions. For many of the questions This Roadmap is an attempt to bring this exciting and
of relevance to HEP research, quantum simulating the yet challenging area of research to the spotlight, and
Standard Model (SM) and particularly the strong force is to elaborate on what the promises, requirements, chal-
a primary objective, but simulating a range of effective lenges, and potential solutions are over the next decade
field theories (EFTs), conformal field theories, or proto- and beyond. Figure 1 presents a schematic overview of
type models of quantum gravity, will also be a critical the aspects of quantum simulation program for HEP over
component of the program for their applications in low- the next decade that are studied in this Roadmap. The
energy and high-energy regimes in nature. Finally, given abundant Appendices contain more in-depth discussions
the underlying simulation to be performed, the proper on several topics along with relevant references. It should
choice of the quantum simulator becomes essential and, be noted that this document concerns only research direc-
importantly, it will likely be the case that special-purpose tions that directly impact the mission of the field of HEP as
quantum simulators for HEP will be required. Such ded- defined within the U.S., that is to “explore the fundamental
icated simulators may come in different varieties given constituents of matter and energy, and to reveal the pro-
different existing physical architectures, and their develop- found connections underlying everything we see, including
ment can be facilitated by a co-design process involving the smallest and the largest structures in the universe” [1].
HEP scientists and quantum scientists and developers. The Roadmap, therefore, does not discuss problems of

027001-3
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

relevance in other closely related fields, such as nuclear important ingredients to collider observables. Such non-
and hadronic physics [2] unless progress in those prob- perturbative ingredients often arise in EFT approaches,
lems are expected to impact theoretical and experimental in which short-distance physics, which is calculable
research in HEP. in perturbation theory, is separated from long-distance
physics, which is often nonperturbative. One of the well-
known examples of such nonperturbative quantities is the
parton distribution function, but other ingredients such as
II. PHYSICS DRIVES
jet and soft functions arise, for example, in soft collinear
HEP is full of important theoretical questions whose effective theory (SCET) [15–18].
answers seem to be intractable using traditional classical- Quantum computing is also required to understand
computing techniques. Addressing these problems, which strongly coupled matter at high density or far from equi-
span many different subfields, including collider physics, librium. In the coming decade, results from heavy-ion
neutrino (astro)physics, cosmology and early universe and proton collisions at the Relativistic Heavy Ion Col-
physics, and quantum gravity, will have a major impact lider (RHIC) and the Large Hadron Collider (LHC), as
on our understating of nature and its underlying work- well as gravitational-wave searches from the Laser Inter-
ing principles at the most fundamental level. High-energy ferometer Gravitational-Wave Observatory (LIGO), will
physicists are actively identifying concrete areas that quan- lead to an unprecedented level of experimental data prob-
tum simulation methods and technologies can advance in ing strongly interacting matter. These problems are again
the coming decade. inaccessible to standard Monte Carlo lattice-gauge-theory
Probing the SM at the highest possible energies is usu- techniques due to a sign problem. Quantum computers
ally achieved using high-energy particle colliders, and par- offer the promise of reducing the computational complex-
ticle collisions are notoriously difficult to describe. Strong ity in this problem from exponential to polynomial by
interaction, being asymptotically free [3,4], is described naturally incorporating entanglement as quantum states
in terms of weakly interacting quarks and gluons at high [19,20]. Such simulations should lead to precision theoret-
energies. However, since the final and, in many cases, ical results for the QCD equation of state and the behavior
initial, states of the collisions involve hadrons, nonpertur- of phase transitions in strongly interacting matter [21].
bative physics is required for any full description of the The ability of quantum simulations to perform real-time
events. Since the simulation of collider physics requires evolution will also provide insight into the approach to
understanding real-time dynamics, normal lattice-field- equilibrium in strongly interacting systems [22]. There is
theory techniques based on a Monte Carlo approach is tantalizing evidence from heavy-ion and proton-collision
hampered by sign problems [5,6]. For this reason, parti- experiments at RHIC and LHC that the strongly interact-
cle collisions are nowadays described theoretically using ing matter thermalizes in a remarkably short period of
various types of approximations [7,8]. Processes at the time, in disagreement with naïve scaling arguments. One
highest energies are computed using a perturbative eval- possible solution to this puzzle is that the dynamics of
uation of the full quantum mechanical amplitudes [9]. quantum entanglement plays a role in the equilibration pro-
The production of a large number of additional partons cess [23–26]. Furthermore, it is expected that entanglement
is traditionally described by a parton-shower algorithm Hamiltonians and associated entanglement spectrum [27]
[10–12], which is based on classical emission probabilities. can reveal unique properties of the quantum many-body
Finally, phenomenological models are used to describe system such as whether and how the system thermal-
how the resulting partons hadronize to form color-neutral izes and if phase transitions occur [28–31]. Studying such
hadrons. Quantum computers hold the promise to simulate behavior in large, far-from-equilibrium strongly interact-
scattering processes from first principles, and, in princi- ing quantum matter is prohibitively difficult with classical
ple, without any uncontrolled approximations. The basic computers. Quantum simulators represent a natural tool to
idea is to approximate the continuous model by a set of create such conditions and track the real-time evolution of
discretized formulations, which can systematically approx- entanglement in strongly interacting matter.
imate the original model with increased precision. It is Another area of HEP where classically intractable prob-
believed that all relevant ingredients required to compute lems are present is neutrino (astro)physics. A prime exam-
the S matrix can be calculated on a quantum computer ple is core-collapse supernovae and neutron-star mergers,
using resources that scale only polynomially with the where the very large number of neutrinos present require
number of lattice sites [13]. This was demonstrated explic- taking neutrino-neutrino interactions, both within the SM
itly for a scalar field theory [14], but is believed to also and beyond, into account. Collective neutrino oscillations
remain true for more complicated field theories, such as have been shown to potentially have an important impact
the gauge theories of the SM. Understanding the behavior in supernova environment, both in the neutrino-driven
of the strong interaction in the nonperturbative regime can explosion mechanism and in the ensuing nucleosynthe-
allow the calculation of long-distance quantities that are sis in the ejected material [32–34]. This results in a full

027001-4
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

many-body problem, hence an exact solution to the dynam- effective potential, but the phase-transition history defined
ical evolution of flavor from a general initial configuration solely by the conventional critical-temperature calculation
amounts to a computational cost, which is exponential in can be misleading [68]. As a result, it is necessary to
the number of neutrinos involved. Quantum computing can perform more detailed calculations of the nucleation mech-
be used to simulate the propagation of neutrinos, in partic- anism. A first priority for quantum research in this area is
ular collective neutrino oscillations, with resources that are to rigorously define matrix elements for bubble properties
polynomial in the number of neutrinos. Due to the simple and exploratory calculations of them [52–54]. Addition-
structure of the neutrino-neutrino Hamiltonian, it is also ally, the interplay between phase transitions and curved
possible, at least for simple geometries and energy distri- spacetime can be investigated with the noisy intermediate-
butions, to map neutrino systems directly into the degrees scale quantum (NISQ) simulators for low-dimensional toy
of freedom and interactions of quantum simulators [35,36]. models [55–58]. Another mystery of the early universe
In a rather different direction, the measurement of fun- is what accounts for around 85% of the matter observed
damental properties of neutrinos, such as their absolute only via its gravitational effects. Light dark matter such
masses, mixing angles, and the CP-violation parameters as axions typically require a more complex and nonequi-
depends upon accurate determination of neutrino cross sec- librium history to be produced. To extract the axion-mass
tions with the material, often nuclear isotopes, used in the information, QCD free energy as a function of the CP-
detector in experiments such as in the Deep Underground violating phase and temperature has to be calculated, a
Neutrino Experiment (DUNE). A full description of the problem that is hindered by a sign problem classically.
neutrino-nucleus scattering cross section is a formidable With real-time simulations of the early universe, quantum
theoretical challenge due to the wide energy range required computers can also simulate the nonequilibrium histories
for the analysis of the experiment. Future quantum sim- of various types of dark matter, such as light dark matter,
ulations can help these efforts by allowing for both topological defects [69], and primordial black holes [70].
more efficient representations of the nuclear target’s wave Besides the important phenomenological questions
function, and the extraction of inclusive and semiexclu- raised above, constructing a complete and convincing
sive reaction cross sections with controllable uncertainties quantum theory of gravity is a grand challenge facing fun-
[37,38]. Early attempts with current digital quantum damental physics. The main goals are to develop a deeper
devices are encouraging (see, e.g., Refs. [38–42]) but to understanding of the fundamental laws of nature, to resolve
achieve the required accuracy, large-scale error-corrected long-standing puzzles about what happens inside black
quantum simulators are likely required. holes, and to explain the initial conditions in the history of
Since the universe is inherently quantum, predictions the universe. Despite limited guidance from experiments,
about earlier epochs of the universe may need to include remarkable progress is being achieved, examples being
quantum effects. Multiple phenomena across cosmology the anti-de Sitter and/or conformal field theory (AdS/CFT)
and the early universe, including inflation, baryon asym- duality [71], augmented by the realization that bulk geom-
metry, phase transitions, and dark matter, may need to be etry emerges from boundary entanglement [72]. Despite
described by the nonequilibrium dynamics of nonpertur- this progress by melding insights from the holographic cor-
bative quantum fields, which again is not possible using respondence with ideas from QIS [73–88], much is still
standard lattice-field-theory methods and requires approx- missing from our current understanding of quantum grav-
imations that lead to uncontrolled systematics. Simulat- ity. Thanks to holographic duality, bulk gravitational phe-
ing such phenomena ab initio provides opportunities for nomena can be described by a “dual” boundary quantum
demonstrating practical quantum advantage. The inflation- system consisting of many particles strongly interacting
ary phase is terminated by nonperturbatively transferring with one another. In principle, this boundary system can
energy to particle degrees of freedom through reheating be simulated using a quantum computer, opening oppor-
and preheating. These far-from-equilibrium and nonper- tunities for exploring quantum gravity in table-top exper-
turbative processes could leave imprints on the sky today iments. One needs to find a (nongravitational) quantum
[43–45]. Though in the long term, large-scale quantum system that has a gravitational dual, realize this quantum
simulations of quantum inflationary fields are desired [46– system in a feasible experiment, and develop a dictionary
49], near-term studies can be useful as well [46,50–64]. relating gravitational phenomena of interest to measur-
Nonequilibrium dynamics are also required for generating able observables in the quantum system. More tractable,
the observed baryon asymmetry [65]. Standard treatments but still ambitious, targets with good gravitational duals
rely upon the dynamics when the interaction rate is close would be certain matrix models [89–91]. The information
to the Hubble rate, which is assuming nearly equilibrium of the gravitational geometry is encoded in matrix degrees
behavior [66], but simplistic comparisons can be insuf- of freedom [89,92,93]. Although Monte Carlo simulations
ficient to ensure an adiabatic evolution [67]. Serving as on classical computers provided nontrivial test of this con-
inputs to the time evolution of classical fields, the prop- jecture (see, e.g., Refs. [94,95]), the details of the encoding
erties of nucleating bubbles are usually extracted from an of geometry into matrices have been out of reach. Quantum

027001-5
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

simulation would provide a practical tool in this problem. in higher-dimensional simulations). Quantum simulation
Though spectacular insights into quantum gravity may is a natural mechanism to implement Hamiltonian sim-
not be expected in the next 10 years, the community can ulation of QFTs, however, similar to the development
develop tools, methods, and insights that will lay founda- of the conventional lattice-field-theory program, it will
tions for quantum technology, enabling profound advances need a dedicated program of continuous research and
in fundamental physics. innovation in theory, algorithm, hardware implementa-
While the physics drives of the program as enumerated tion, and co-design to reach the level of maturity that
in this Roadmap are the most immediate motivations for is needed for addressing the physics drives of the HEP
using quantum technologies in simulation over the coming program.
decades, it is likely that as high-energy physicists discover The Hilbert space of QFTs is infinite dimensional as
the power of quantum simulation, and enjoy access to the fields are defined on an infinite and continuum spacetime.
current and next generations of quantum simulators, new For bosonic theories, the local Hilbert space of the fields is
problems and additional physics goals will be identified also infinite dimensional. Simulating QFTs on any finite-
and explored. capacity quantum hardware requires truncating the extent
of the volume, discretizing the space, and digitizing and/or
truncating the on-site Hilbert space of the bosons. In the
III. UNDERLYING SIMULATIONS
context of Hamiltonian simulation, various approaches are
Quantum field theories are the underlying mathemati- proposed to make QFTs finite dimensional in a system-
cal description of three of the four fundamental forces in atic way. Furthermore, evolution in a quantum simulator
nature, constituting the SM of particle physics. They are occurs in a digital or discrete manner leading to time-
also closely connected to theories of quantum gravity via digitization errors or in an analog or continuous manner but
the bulk/boundary duality. QFTs in form of EFTs also pro- with potentially approximate Hamiltonians implemented.
vide a systematic mechanism to organize interactions in As with the development of the conventional lattice-field-
nature at low energies where there exists a separation of theory program, quantifying the systematic uncertainties
scales, and coarse-grained descriptions of degrees of free- in quantum simulation of lattice field theories, includ-
dom are applicable, as well as high energies where the ing renormalization and continuum limit, finite boundary
unknown physics beyond the SM can be introduced via effects, boson-field digitization and truncation errors, and
higher-dimension operators of SM fields. Various pertur- inaccuracy in the time-evolution stage is crucial to achiev-
bative and nonperturbative approaches to solving QFTs ing realistic results and resource estimates. Effective field
have been developed and successfully applied over the theories may provide a pathway to such characterizations
years. In the realm of SM predictions in the nonperturba- as with the conventional program. While initial efforts
tive regime of the strong interactions, lattice-field-theory along these lines have started (see Appendix F), these
methods have proven the most reliable approach to date, questions will continue to be at the center of theoretical
providing critical input to many experimental programs in studies in the upcoming years.
HEP, see, e.g., Refs. [96–103] for recent reviews. Nonethe- The basis chosen to span the Hilbert space informs the
less, the conventional lattice-field-theory program based truncation scheme, and proper choices can result in faster
on the path-integral formalism of QFTs relies on Monte convergence to the desired asymptotic limits and/or lower
Carlo importance-sampling methods, and this statistical computational cost, classically or quantumly. For gauge
feature has halted progress in several problems. These theories, the manifestation of gauge invariance and of the
include finite-density systems where a signal-to-noise or local Gauss-law constraints are another determining fac-
a fermionic sign problem demands an exponential increase tor in adopting the proper formulation. The Hamiltonian
in generated samples and measurements with the system formulation of lattice gauge theories (LGTs) provides the
size. The same issue has prevented meaningful progress starting point for a range of frameworks: (i) irreducible
in real-time problems such as scattering processes, except representation (irrep) or electric-field bases that diagonal-
for those at low energies and low inelasticities that can ize the electric Hamiltonian, which are a natural basis for
be addressed by indirect methods [104–109]. That is expressing the Gauss laws, and can be truncated to a finite
because the gauge-field importance sampling is enabled by number of irreps without breaking gauge invariance [114–
a Wick rotation to Euclidean spacetime. A Hamiltonian- 118], (ii) prepotential and loop-string-hadron bases which
simulation approach does not encounter such issues but build the Hilbert space that satisfies the non-Abelian Gauss
the size of the required Hilbert space scales exponen- laws via constructing gauge-invariant operators locally,
tially with the system size. This has limited classical- and impose only an Abelian constraint on the link [119–
computing methods, including tensor-network approaches 123], (iii) group-element bases that simplify the implemen-
that have so far mostly targeted simpler low-dimensional tation of gauge-matter and the magnetic Hamiltonians, and
theories and in systems without volume-law entanglement are a more economical choice in the weak-coupling regime
[110,111] (see Refs. [112,113] for examples of progress relevant for approaching the continuum limit [124,125],

027001-6
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

(iv) dual or magnetic basis, which diagonalizes the mag- Digital quantum algorithms for QFTs for both near- and
netic Hamiltonian, is obtained via discrete group Fourier far-term era of quantum computing need to be developed
transforms when applicable, or via tensor-renormalization in the context of various Hamiltonian formulations and the
group methods in given systems [126–132], (v) light-cone particular quantities that are aimed to be extracted. Near-
quantization of fields leading to basis states in the light- term era refers to at least the next decade of hardware
cone momentum occupation-number basis with encoding- developments, where the qubit resources are scarce and the
cost benefits over other formulations but with subtleties gate fidelities, particularly those of the entangling gates,
regarding the zero modes and UV-IR mixing during renor- are not high enough to enable reliable error-corrected
malization [133–135], (vi) quantum link model [136,137] simulations. In the far-term era, when fault tolerance is
or qubitization [138–140] formulations of gauge theories achieved, auxiliary qubit registers can be used at ease but
that assume a finite-dimensional on-site Hilbert space for error-corrected synthesis of non-Clifford single-qubit gates
the gauge fields, but introduce an extra dimension that such as T gates becomes costly. Each scenario, therefore,
needs to be taken to infinity, or alternatively that take requires different resource optimizations. For non-Abelian
advantage of universality classes near the critical points to gauge theories, for example, the noncommuting algebra of
approach the continuum limit without the need to introduce the group generally introduces costly arithmetic in terms
the extra dimension, (vii) matrix models, which provide a of both near- and far-term resource measures. Addition-
way to map gauge theories to quantum mechanical mod- ally, the choice of simulation algorithms that amounts to
els via dimensional reduction, while preserving some of the method of approximating the unitary time evolution, as
the interesting nonperturbative dynamics and structure of well as qubit encoding that amounts to storing fermionic
the parent QFT [89,141–144], among a few other choices. and bosonic quantum numbers (and the Fermi statistics
We do not know what the most optimal formulation is of fermions), greatly impacts resource requirements of the
for QCD, that is a SU(3) LGT in 3+1 D is with multi- simulation, and full or partial retainment of symmetries
ple flavors of quarks in the fundamental representation, as and conservation laws at each stage of the evolution. While
this theory has not yet been fully studied in the context algorithmic progress for Abelian and non-Abelian gauge
of the formulations mentioned above. The next decade of theories of relevance to the SM has been significant in
developments will shed light on this important theoretical recent years [117,118,163,164], as with classical comput-
question. ing, the most efficient algorithms are likely not going to be
An important ongoing investigation is to what extent the early algorithms. Furthermore, the asymptotic scaling
preserving symmetries, and, particularly, local gauge sym- with simulation accuracy must be greatly improved in the
metries and the Gauss law, matter in a quantum simu- coming years by taking advantage of physics inputs [165–
lator. Questions regarding the robustness of simulations 169] and insights from empirical analyses [162,170,171].
to gauge-symmetry-breaking perturbations in the dynam- The algorithms necessary for preparation of nontrivial ini-
ics are starting to be explored in simpler gauge theories tial states such as hadrons and nuclei, quantities such as
and quantum link models [145,146]. Furthermore, sev- scattering and transition amplitudes, structure functions,
eral proposals and algorithms are put forward in recent equal- and out-of-time correlation functions, and entan-
years for detecting and discarding the Gauss-law violations glement measures, as well as efficient state-tomography
[147,148], and for suppressing coherent gauge-symmetry- methods, need to be also advanced. Furthermore, proper
violating noise [118,149–161], taking advantage of fea- definitions of entanglement in the context of lattice gauge
tures like introduction of energy penalties, classical noise theories (which represent nonseparable Hilbert spaces due
and Zeno effect, quantum control, dynamical decoupling, to local constraints) must be developed within various
random rotations of the state throughout the evolution via Hamiltonian formulations [172–175].
unitaries generated by the symmetry (or pseudosymme- Analog quantum simulators provide another potentially
try) operator, and controlled operations in digital circuits. promising path to quantum simulation of QFTs. Ana-
The value of these strategies and their limitations must be log simulators may naturally offer fermionic and bosonic
confirmed in realistic experiments [162] and in the con- degrees of freedom removing the need for expensive qubit
text of each of the Hamiltonian formulations discussed encodings, or provide tunable interactions between larger
above. Furthermore, more targeted strategies need to be local Hilbert spaces like qudits. On the other hand, one
developed to suppress the incoherent noise and, more gen- needs to engineer the interactions of these degrees of free-
erally, to represent gauge-theory simulations as dynamics dom to represent the dynamics of the QFT Hamiltonian of
of open quantum systems coupled to the environment. In interest. Analog simulators may come in larger sizes with
the fault-tolerant era, while the hardware-based errors are sizable Hilbert spaces and with two- and three-dimensional
substantially reduced and can be corrected, the errors asso- geometries, and offer the possibility of engineering inter-
ciated with inexact algorithms persist. Nonetheless, these action graphs representing curved spacetime [176,177],
can always be bounded and reduced systematically by hence enabling studies pertinent to black-hole physics and
increasing computing resources. quantum information [58,178]. These exciting possibilities

027001-7
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

are faced with the challenge of finding the proper plat- methods (as long as sign and signal-to-noise problems
form for a given problem both in terms of similarities of are not encountered) or tensor-network-inspired meth-
Hilbert spaces, the mapping of the symmetries, and the ods (as long as states are area-law entangled). Quan-
ability to engineer complex interactions with the knobs tum computers can then use this initial-state input to
available in the simulator. For example, simulating the perform the classically intractable time evolution, hence
dynamics of both Abelian and non-Abelian gauge theories saving quantum resources needed for state preparation
in 2+1 and higher dimensions have proven hard [179–184] [207,208]. Furthermore, various costly stages of a con-
given the higher-body interactions present when working ventional lattice-field-theory calculation may be expe-
in the irrep basis, and nonlocal interactions when working dited by quantum processors. These include generation
in dual basis. Some progress has been reported in recent of gauge-field configurations, matrix inversions [209],
years, but a first implementation of complete building Wick contractions, and design of suitable interpolating-
blocks of a lattice gauge theory with high fidelity remains operator sets for the states [210]. Over the next decade,
a critical goal of the program in the coming decade. If the community will need to continue to identify and
such simulations become possible and can be scaled, one advance such hybrid classical-quantum approaches to
needs to face the question of rigorous error-bound analysis, simulation.
error corrections, and verifiability that are less developed The framework discussed so far can be used to sim-
for analog simulations compared with their digital coun- ulate the full S matrix of an underlying QFT. However,
terparts. In the same spirit, continuous-variable quantum in the case of QCD theory, quantities at short distances
computing presents another opportunity for quantum sim- (high energies) can, in most cases, be reliably calculated
ulating QFTs, as it may provide a more natural way of perturbatively, and it is the physics at long distances (low
treating bosonic fields in the simulation [185,186]. energies) that is nonperturbative in nature. EFTs can be
It is important to emphasize the role of hardware used to separate the physics at long and short distances,
implementation and benchmarks in guiding the course of with the EFT describing the long-distance dynamics, while
developments in quantum simulation. Many recent exper- short-distance effects can be included through coefficients
iments and implementations of a variety of QFT problems that are obtained by perturbative renormalization-group
on a range of quantum architecture in both analog and matching, via nonperturbative methods, or by matching
digital modes (see, e.g., Refs. [40,115,116,118,162,187– to experiment in regimes where the EFTs are applicable.
198]) have generated an exciting platform for communica- The EFT setup then defines the long-distance ingredients
tions and collaborations with leading experts in quantum that can, in principle, be computed on a quantum com-
hardware and software technology. It has also generated puter, together with a prescription of how to combine it
proposals and experiments dedicated to developing sim- with the short-distance physics to obtain a physical observ-
ulators suitable for QFTs on a range of platforms, see, able. This strategy saves quantum resources as simulating
e.g., Refs. [35,150–152,199–205]. Continuing this trend the full underlying theory is often costly.
is essential as otherwise theory and algorithmic develop- A well-known nonperturbative quantity in hadronic
ments will be disconnected from the reality of hardware. physics is the parton distribution function (PDF). It
Furthermore, implementations on various platforms assist involves a matrix element of two quark fields separated in a
in verifying the outcome of quantum simulation once the lightlike direction, which can not be calculated using tradi-
limits of classical computing are challenged in the forth- tional lattice-field-theory techniques directly due to a sign
coming simulations. To achieve meaningful progress, over problem. While several techniques have been put forward
the next decade a series of models from low-dimensional to make calculations of PDFs possible with notable success
theories and simpler gauge groups need to be identified and [211–215], quantum computers give rise to the possibil-
progressively made more complex to follow, or ideally, ity to compute the matrix element relevant for the PDFs
guide, hardware developments. directly and from first principles. Several proposals have
Finally, hybrid classical-quantum approaches to quan- been put forward in recent years to demonstrate how PDFs
tum simulation should be taken advantage of to a full can be accessed on a quantum computer [216–220]. One
degree in both near and far terms. On one hand, variational- can calculate the hadronic tensor on a quantum computer
based algorithms can be proven useful in spectroscopy and then extract the PDF using perturbative information
studies as they rely on classical optimizers to find the for the partonic scattering [216]. Alternatively, the Wilson
ground state of a given Hamiltonian given a nontriv- line can be explicitly constructed using plaquette opera-
ial wave function prepared on a quantum computer, see tors or fermion hopping terms, allowing for an estimate
Refs. [40,115,188,190] for several gauge-theory exam- of the full quantum computation of the PDF [216,217].
ples. They can also be taken advantage of in scat- Finally, a PDF calculation in the NJL model was per-
tering problems as recently proposed [206]. On the formed in Ref. [218] using a variational ansatz for the
other hand, classical computers may prepare nontrivial proton state, and following Ref. [221] for the correlation
states through the use of conventional lattice-field-theory function. At this stage, it is not clear what the realistic

027001-8
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

computational-resource requirements are for computing ported to the nuclear simulations. A few main differ-
PDFs and hadronic tensor to given accuracy, as the com- ences are the presence of additional fermionic species,
plete algorithms, including that needed for preparation of the presence of three- and higher-nucleon interactions,
hadronic states in QCD on a quantum computer, are either and the presence of long-range pion-exchange interactions.
nonexisting or premature. Over the next decade, theoretical Progress has been made in applying variational techniques
and algorithmic research will improve upon these initial to obtaining the ground-state properties of light nuclei
analyses. [39,40,229], and in algorithms for nuclear-reaction cross
An EFT describing the collinear and soft physics in jet- sections and neutrino-nucleus response functions [37,230].
like collider events is SCET [15–18]. The PDFs mentioned Further progress is needed to optimize and advance such
above are collinear matrix elements in the theory, how- algorithms by applying new developments in simula-
ever soft matrix elements are required as well to make tion algorithms and low-weight fermionic encodings, see
predictions with jets in the final state. Since these soft Appendix G 4. In a similar vein, ab initio many-body
matrix elements are again nonperturbative in nature, quan- calculations based in the underlying few-body hadronic
tum algorithms might be suited to compute them from first interactions are needed to address questions regarding the
principles. This has been studied in a simplified theory, composition of the interior of neutron stars [231–234],
namely a scalar field theory interacting with static sources. which is relevant for the analysis gravitational-wave emis-
All necessary quantum circuits to compute the soft func- sion from merging neutron stars [235]. Such calculations
tion in this example were worked out in Ref. [222], along are computationally challenging, but are crucial in discern-
with necessary quantum circuits and a small-scale simu- ing the role of non-nucleonic degrees of freedom in the
lation on an IBM Quantum device. The progress in this equation of state of neutron stars [236,237]. Quantum com-
problem is correlated with that in simulating lattice gauge puting the many-baryon problem is similar in nature to the
theories in 3+1 dimensions. computations described above for many-nucleon systems.
Another example of important long-distance effects in To constrain the many unknown low-energy constants of
collider events is collinear radiation, which traditionally is the effective hadronic description, it is anticipated that the
described by parton-shower algorithms [7,10,12,223,224]. lattice-QCD program in the few-hadron sector could be
The very nature of a probabilistic Markov-chain algorithm matched to a nuclear and hypernuclear structure program
makes including quantum-interference effects challenging, [40], similar to what is currently promoted using conven-
since collider events typically contain a very large number tional tools [238–240]. The output of such efforts will
of final-state particles. Quantum interference effects that subsequently impact research in HEP, including problems
can be present are effects at subleading orders in the inverse in the intensity and cosmic frontiers.
number of colors, and interference between amplitudes
with different intermediate particles and different internal
IV. SIMULATOR REQUIREMENTS
kinematics. This problem, therefore, is a suitable candidate
for quantum simulation. A quantum algorithm has been Currently, quantum hardware is being developed in a
developed in Ref. [225] that reproduces the regular parton variety of forms. This includes, on one end of the univer-
shower, while by computing all possible amplitudes at the sality spectrum, special-purpose simulators whose intrinsic
same time, it also includes quantum interference effects. or engineered Hamiltonian emulates that of the simulated
More work is required to develop a full parton-shower theory, and on the opposite end, gate-based universal sys-
algorithm for the SM, and to find ways to include the tems with, in principle, the ability to accurately implement
most relevant quantum interference effects, such as color any unitary operation. The choice of the suitable archi-
interference. tecture, and the mode of operation of the simulator given
Finally, low-energy EFTs of nuclear interactions are the simulation problem, impacts the efficiency of the algo-
important for the HEP mission as they provide a con- rithms, and the accuracy achieved with finite resources. It
sistent framework to describe the nuclear targets used in is important to understand the capabilities and limitations
high-energy experiments such as in long-baseline neu- of current leading simulating hardware and their prospects
trino experiments, neutrinoless double-β decay exper- for simulating HEP models, how peculiarities of the NISQ
iments [226], and direct dark-matter detections [227, era of computing may provide exploratory opportunities
228]. Dynamical properties of nuclei like inelastic cross and open the door to potential co-design efforts, and what
sections are challenging to compute classically, espe- form of software and compiler developments is required
cially for semiexclusive scattering in medium- and large- for HEP applications in the NISQ era and beyond.
mass nuclei, and quantum simulations have the poten- Today, functional analog quantum simulators exist
tial of being impactful in such problems. Simulation of based on atomic, molecular, optical, and solid-state tech-
these theories is similar to a quantum chemistry sim- nologies, and are being constantly optimized for pro-
ulation, given the nonrelativistic nature of interactions, grammable quantum simulation. Systems of neutral atoms
thus techniques developed in quantum chemistry can be have been used as analog quantum simulators since the

027001-9
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

first production of quantum degenerate Bose and Fermi simulators are, unfortunately, still far away from present-
gases [241–244]. They nowadays come in the form of ing the essential capabilities for simulating gauge theories
purpose-built optical lattices with the addition of quantum of the SM. A hybrid approach can be explored, where
gas microscope [245–247], allowing very large systems of native and more versatile sets of operations are used,
neutral atoms to be placed in a single many-body state. but the evolution is digitized to avoid the need for the
These systems can be made in various configurations, challenging task of simultaneously applying an increas-
with a range of optical lattices with controllable geometry, ingly large number of terms in the local Hamiltonian of
dynamical couplings, and Floquet engineering [248] avail- gauge-field theories. Finally, while trapped-ion and cold-
able. Furthermore, cold atoms held in independently mov- atom platforms have been explored more extensively in the
able optical tweezers and driven by laser light into Rydberg context of quantum simulation of gauge theories, it is not
states have been developed as a platform for quantum yet known if a range of other platforms will provide unique
information processing, with successful demonstration of opportunities for this task. This is a question that only com-
simulating classically intractable quantum many-body sys- bined theoretical and experimental research can illuminate
tems [249,250]. Trapped-ion systems are also among the in the coming decade.
popular quantum-simulation platforms in which atomic Digital quantum computing implements algorithms as
ions, that are naturally interacting via long-range Coulomb sequences of universal gates on the underlying qubit archi-
interactions, are confined through electromagnetic fields, tecture, based on superconducting systems, cavity QED,
forming one-dimensional (1D) or two-dimensional (2D) neutral atoms, or trapped ions, among other technolo-
crystals in space. Each atomic ion stores an effective spin gies. Current technologies are limited in scale below the
with a long idle coherence time. A tunable long-range 100-qubit level but roadmaps exist for scaling up these
Ising interaction between all ion pairs is achieved via cou- system considerably. Trapped-ion quantum computers fea-
pling off resonantly the spin and the motional degrees ture qubits with long coherence times (minutes) [258],
of freedom, a feature that had led to many interesting single-qubit with > 99.9999% average fidelity [259] and
quantum simulation experiments of spin systems with two-qubit gates with > 98% fidelities [260,261], and
long-range interactions [251]. Trapped-ion systems can reported readout fidelity of approximately 99.97% [262].
be utilized to simulate three- and multispin Hamiltoni- Ion chains with tens of functional qubits are achieved in
ans as well [150,199,252,253]. They can also operate with current systems, and importantly, are all mutually con-
qudits where extra internal atomic levels are addressed nected. This feature greatly reduces circuit depths as the
[254,255]. Furthermore, analog and hybrid analog-digital qubits do not need to be swapped and placed in proximity
implementations are plausible using phonons as dynamical of each other to enable a gate between them. The con-
degrees of freedom [201,202,256]. Among other notable trol of large ion chains are limited by the dense motional
platforms [257] with potential prospect for analog simula- (phonon) modes, demanding that ion crystals be bro-
tion of models of relevance to HEP are laser-cooled polar ken into spatially separated modules and connected via
molecules, cavity quantum electrodynamics, supercon- either photonic links [263,264] or shuttling [260]. Another
ducting quantum circuits, and dopants in semiconductors popular digital quantum computing platform is super-
such as in silicon, each presenting unique opportunities but conducting electronics-based quantum processing units
are currently restricted in scalability, controllablity, and/or (SC QPUs). Being semiconductor systems, SC QPUs can
coherence times, see Appendix H. leverage extremely high-purity solid-state materials and
HEP simulations, particularly those rooted in QFTs, sophisticated material-processing techniques to produce
demand the development of simulators of large scale that QPUs with coherence times approximately 100’s µs, high
exhibit good quantum coherence and high-quality read- single-qubit, 99.8–99.96%, and two-qubit, 98%–99.6%,
out, but perhaps more importantly, significant amount of gate fidelities [265], and to realize chips of varied qubit
control beyond what has been customary in the quantum counts from a few to close to 100. Imperfections in mate-
simulation of simpler spin systems in the past. Since the rials, control systems, QPU design, and electromagnetic
analog simulator and target system often do not share the environment are among the culprits of QPU operation per-
same fundamental symmetries or connectivity of interac- formance—areas that will be constantly improved. Other
tions, computing resources are inevitably lost to encoding emerging platforms, such as highly scalable neutral atoms,
overheads. Additionally, these restrictions might make it are likely to introduce another candidate architecture for
difficult or impossible to engineer certain Hamiltonians. digital quantum computing over the next decades.
With simulators exhibiting fermionic and bosonic degrees New programming paradigms, compilation, and tran-
of freedom, and newer modalities involving efficient mul- scription processes will be necessary to understand how
tiqubit operations, new opportunities to mitigate these best to utilize digital quantum computers in HEP applica-
overheads will likely arise, and the issues with encod- tions, given their enormous qubit and gate requirements,
ing fermionic statistics and the need for low truncation of see, e.g., initial estimates in Refs. [14,117,118,163,164].
bosonic modes may be circumvented. The existing analog For the purpose of simulating gauge theories of the SM,

027001-10
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

for example, it may be best to work with the formu- number of quantum-simulation problems of importance
lations and encodings that retain the locality of inter- to HEP [280,281]. Another approach involves increasing
actions, both among bosons and among fermions and the size of the operational Hilbert space by using three-
bosons, depending on the connectivity pattern inherent and higher-level systems in place of qubits. On another
to the hardware architecture used, in order to minimize front, in the NISQ era of computing, one could encode
costly swap operations. For the purpose of quantum sim- the portion the problem that is harder to compute with
ulation, any model Hamiltonian with local and semilo- classical methods on a quantum processor (e.g., real-time
cal interactions can be decomposed into smaller units of dynamics), and combine the results when possible with the
time evolution through the Trotter-Suzuki expansion [266] components that can be evaluated with more ease on a clas-
and efficiently decomposed into universal gate operations. sical computer (e.g., preparation of certain states). Finally,
However, other simulation algorithms that are costly for the translation of a quantum algorithm into specific gates
present-day hardware may be of value in the fault-tolerant must be done with an eye on minimizing quantum-resource
and large-scale era of quantum computing for HEP. While requirements in the NISQ era. For example, the original
proof-of-concept studies of error correction with differ- quantum algorithm for computing scattering amplitudes
ent encodings have been completed on various platforms in a scalar field theory [13,14] is costly for present-day
[267–272], no system to date has sufficient resources to quantum computers, but several improvements have been
meaningfully utilize it in real-world algorithms. Through suggested since then to bring down the cost considerably
deep user interaction, co-design, algorithm innovation, [282–284]. Furthermore, strategies for fully or partially
and continued improvement in digital-computing hard- removing the redundant gauge degrees of freedom may
ware, the field is moving toward accelerating the timeline reduce the encoding overhead, but such formulations can
toward universal fault-tolerant quantum computation. As change the local nature of interactions inherent in the
such, HEP physicists need to keep an eye on developing original theory and may result in higher gate complexity
optimal fault-tolerant algorithms, and study the interplay [129,162,187,285]. As a result, careful analysis of quan-
between error-correction protocols and gauge constraints tum resource costs and the susceptibility to noise will be
in gauge-theory simulations [273]. essential in simulating gauge theories in the NISQ era.
NISQ era refers to an era of quantum computing where Error mitigation is another important ingredient to make
noisy non-error-corrected operations are performed on sense of the measurements in the NISQ era. There are two
devices with approximately 50–100 qubits [274]. It is general types of errors that can affect a quantum simula-
imperative to not dismiss the possibilities provided by the tion: readout errors and the gate errors. Currently, readout
NISQ hardware, and to take advantage of available devices errors have error rates between < 1 to 10%. The largest
with various capabilities and capacities, see Appendix J. gate errors occur in entangling operators, such as the CNOT
Developing, optimizing, and testing the building blocks of gate, with a typical error rate slightly below 1%. However,
QFT simulations from simpler (Abelian) low-dimensional gate errors accumulate and become dominant for longer
models to more complex (non-Abelian) models in higher circuits. One way the readout errors can be mitigated is
dimensions is a near-term task that can be started with by preparing a given state and recording the measure-
existing NISQ devices [40,115,116,118,162,187–191]. In ment of this state, hence providing a conversion matrix
this context, error mitigation, device comparisons, and to apply to any other measurement, a technique that can
benchmarking are essential for making progress. The HEP take advantage of developments in collider physics [286].
community will continue to develop and apply new NISQ- Gate errors include stochastic and coherent errors. Stochas-
tailored simulation algorithms in the coming years, and to tic errors can be understood as either coherent errors with
determine if any quantum advantage can be anticipated in randomly varying control parameters or as processes that
the HEP applications in the NISQ era. entangle the system with its environment. Coherent errors,
There are a few promising approaches that can produce i.e., those arising from collective couplings and includ-
meaningful results even with shallow circuits and without ing unitary noise, can be turned into incoherent errors,
active error-correction sequences in the NISQ era. These i.e., those having a stochastic and hence nonunitary nature,
include hybrid algorithms, such as variational quantum via certain methods [287–290]. A common method to
eigensolver (VQE) [275,276] and quantum approximate reduce stochastic gate errors is zero-noise extrapolation,
optimization algorithms (QAOAs) [277], that divide clas- which measures a given circuit at different noise levels
sical and quantum resources such that only steps that (by changing the entangling-gate operation time or by
require probing a large combinatorial space are executed replacing each gate or unitary block by a larger number
on qubits [278]. Many early demonstrations of these of gates and operations), and then extrapolates it to zero
approaches arose in the context of lattice gauge theories noise [115,288,291–295]. Among other noise-mitigation
[40,115,188,190]. Quantum annealers, that are nonuniver- proposals [296–301], one with special relevance to HEP
sal quantum optimizer machines [279], have also proven applications is to use the symmetries of the simulated
useful in this context, with recent implementation of a system, such as gauge invariance, to detect and mitigate

027001-11
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

errors [148,302,303]. Further noise-suppression tech- take the programs written in high-level programming lan-
niques at the level of the hardware, and noise-mitigation guages to efficient low-level code that can be run efficiently
schemes at the level of algorithms, may emerge from com- on quantum computing hardware. Open-source software
munications and collaborations among the HEP and QIS and tools, such as those that enable debugging and validat-
communities in the coming decade(s), given the high- ing quantum computer results, will be needed. One could
energy physicists’ expertise in suppressing and mitigating also utilize quantum computers as well-controlled analog
errors and background noise in precision experimental set- systems. This requires the users to effectively engage
tings. While error mitigation will be important to lower across the quantum computing stack, and to have access
the levels of noise of NISQ devices, and to obtain more to software layers that generate pulses needed to drive the
accurate results from them, error correction will likely be interactions of relevance to HEP applications.
needed to bring about the true power of quantum com- One of the issues encountered during the adoption of
puters. This is the case especially for deep circuits, as classical-computing resources for solving domain-specific
the number of experimental measurements required for problems, such as lattice-gauge-theory simulations, was
error mitigation grows exponentially with the circuit depth the diversity of paths and paradigms that required exten-
[304–306]. sive effort to bring to a standard and uniform state. Given
Last but not least, there is a need for expanding the the fast-evolving scene of quantum hardware technology,
programmability of quantum computing devices for test- the computing paradigm might substantially shift away
ing new and creative quantum algorithms. A variety of from what may be considered standard today. Further-
low-level programming languages akin to assembler-level more, the diversity of ideas and paths is a necessity at
programming models in classical computing have been such an early stage of developments. Nonetheless, the HEP
developed over the past decade. Transpilers and early community needs to arrive at a point where consensus
low-level compilers are being developed to optimize the is reached on what the best simulation strategies are to
user-generated operations—generally referred to as cir- achieve the stated goals faster, and emphasize those paths
cuits—into a shorter and more efficient set that can be and paradigms collectively so that the QIS community
translated to pulses and other fundamental operations on would recognize the concrete hardware and software needs
quantum hardware. Many of these tools also incorpo- of the HEP problems.
rate hardware-specific knowledge, such as topologies and It is not easy to predict exactly when the quantum simu-
error rates, to deliver the best performance possible. To lator technologies will reach the point where they can make
ease the programmability of quantum computers, higher- a difference in our simulation capabilities in HEP problems
level programming frameworks will be needed. Within beyond classical-computing limits. Many factors will be in
the Department of Energy (DOE), efforts are underway play, including the state of financial and intellectual invest-
in the Office of Advanced Scientific Computing Research ment over the coming decade(s), to ensure steady growth
(ASCR) to develop a software toolkit that includes higher- of many correlated subfields within QIS. Furthermore,
level programming models and the necessary functionality the current estimates for the resource requirements for
to develop domain-specific languages [307]. Not all the performing simulations of relevance to HEP, such as QCD-
capabilities desired by the HEP are currently available, as based studies, are still crude or incomplete, and straight-
discussed in more detail below. forward algorithmic improvements can likely change the
To advance HEP applications, and enable the broader estimates by many orders of magnitudes. These consider-
HEP community to readily participate, models and ations set the priorities for the coming decade: (i) there
approaches for programming quantum computers need to is a pressing need for better algorithms and adoption of
mature to a level in which abstractions and library-based more efficient simulation paradigms such as analog, hybrid
methods can be adopted to expedite programming and analog-digital, and hybrid classical-quantum approaches
ensure portability. This will require the development of that can reduce the time to solution in many problems of
programming languages, potentially domain-specific lan- interest in HEP, and (ii) as the HEP community awaits the
guages, that can readily express the discretization and large-scale fault-tolerant quantum computers, it needs to
complex interactions, such as the need to describe the identify the problems that can still benefit from smaller-
coupling of fermions and bosons and open quantum sys- scale systems that will, nonetheless, be more powerful
tems, provides key building blocks, such as costly function than classical-computing systems in certain problems, and
evaluations using arithmetic subroutines that are preva- qualitative robust results can still be deduced even in the
lent in quantum simulating the dynamics of non-Abelian presence of some level of noise.
lattice gauge theories, or large (sparse) matrix inversion
for expediting Monte-Carlo-based routines present in the
V. QUANTUM ECOSYSTEM
conventional lattice-gauge-theory program. In addition to
programming languages and libraries, the HEP research While it is, in principle, possible to construct gate-based
community would benefit from efficient compilers that can quantum computing algorithms without detailed knowledge

027001-12
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

of the underlying hardware, the resource constraints of resources in HEP can be ensured is the successful USQCD-
today’s systems imply severe limitations on the scale of collaboration model [314]. One could similarly envisage
computations, and imply a need for hardware-aware effi- a meta collaboration of US HEP QIS theorists whose
cient implementations and designs that take advantage main goal would be to strengthen individual efforts, and
of a detailed understanding of system interactions and coordinate access to different hardware platforms available
connectivity, as well as error and decoherence mecha- at national laboratories or commercial sites. It would also
nisms. This favors tightly coupled collaborative models for facilitate communications between researchers developing
executing science programs, in which hardware providers quantum computing hardware and the theorists developing
and domain experts in HEP applications co-develop scien- algorithms and software for HEP applications. Similar to
tific agendas. Among successful examples in the present the USQCD model, such a meta collaboration can bring
day are the DOE-funded Advanced Quantum Testbed up an algorithm-development effort feeding into indepen-
(AQT) [308] at the Lawrence Berkeley National Labora- dent and basically competing individual groups. Direct and
tory (LBNL) and the Superconducting Quantum Materials broad access to leading hardware and software technolo-
and Systems Center (SQMS) at Fermi National Accel- gies, via a community-approved mechanism that can unify,
erator Laboratory (FNAL), where internal and external advocate for, and coordinate the accessibility needs of HEP
collaborators as well as users are engaged in developing scientists may, therefore, be an important component of a
quantum simulators in tight collaboration with scientists future quantum ecosystem in HEP.
developing algorithms. In addition to DOE programs at Since quantum simulation for HEP requires a consid-
the national laboratories, several commercial hardware ser- erably different and much more multidisciplinary skill
vice providers now offer in-house services to help design set than other established HEP areas, collaborations with
efficient hardware-cognizant implementations on demand other disciplines such as atomic-molecular-optical physics,
[309–313]. This will remove the wall between the users solid-state and condensed-matter physics, computer sci-
and cloud-based services with preset features, and will ence, and electrical and material engineering are antici-
allow the needs of the domain scientists to be communi- pated to achieve the QIS goals relevant to the HEP science
cated directly to the hardware and software developers. mission. This is because the types of problems that require
At the university setting, experimental efforts have long quantum simulation in HEP are similar in form to those
been in harmony with theoretical efforts in quantum simu- in the other domain sciences. Furthermore, the underly-
lation at the same or nearby institutions, and such a model ing simulating hardware is itself a physical system, which
will potentially be the key to success in arrangements out- could be taken advantage of in co-designing algorithms
side academia as well. For HEP applications, resources and protocols. It is also the case that the techniques that
and opportunities in co-design efforts will need to be are being developed for the simulation of QFTs may be of
directed toward systematic evaluation of the best methods relevance for quantum sensing and quantum communica-
to simulate Hamiltonians of relevance to HEP efficiently tions. It would, therefore, be useful for funding agencies
and accurately, design and perform state preparation and to investigate ways for funding to flow across the different
tomography in strongly interacting QFTs, understand the domain-sciences areas to flourish essential coordinations
role of decoherence mechanisms, and how best to incor- and collaborations. The formation of the National QIS
porate symmetries and gauge and scale invariance, or take Research Centers [315] has already stimulated interdis-
advantage of them to mitigate or correct the errors. ciplinary research, but funding models that can establish
This co-design requirement also motivates the need for long-term centers with permanent scientists will be impor-
accessible devices that HEP and other domain scientists tant to the continuity of the efforts and completion of
can use to benchmark algorithms and test new ideas at long-term projects. Ensuring a strong HEP involvement in
ease. Therefore, it is important for the HEP community these centers can ensure that the HEP-specific goals are
to identify optimal ways to have access to state-of-the-art being met via the proximity and accessibility of hardware
programmable quantum computers or quantum simulation and software expertise in QIS.
experiments. In the coming years, the ability to access mul- An important aspect of quantum simulation is that there
tiple platforms with a variety of architectures would lead to exists a significant and growing expertise in the private
rapid progress on target problems in HEP, as different plat- sector. In order for HEP to be at the forefront of quantum
forms may be suitable for different problems. Furthermore, simulation, engagement with technology companies is crit-
it is often necessary to run relatively large simulations on ical, and will be mutually beneficial. However, all devel-
classical computers, as numerical tests and benchmarks opments of importance to HEP research from engagements
in the quantum simulation problems are a necessity. It with technology companies need to be future proofed. That
is conceivable that these computations start to require is to say that any advance that enables accomplishing one
nonnegligible time at high-performance computing (HPC) or more objectives of the HEP mission must be able to
facilities, and this requirement should be recognized ahead reside in the community and not be lost behind an IP
of time. One successful model for how accessibility of barrier if a company decides that this line of research is

027001-13
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

no longer a priority. Given the large investment of the anticipated quantum ecosystem, the pipeline for quantum
private sector in the development of quantum comput- education and training should begin even before students
ers, it is important to carefully evaluate how the need enter university. New and creative ways of educating and
of the HEP program is served by the devices that will training junior, midcareer, and senior scientists, engineers,
likely be developed by technology companies. Important and developers need be encouraged, keeping an eye on
considerations are that these devices are developed with an inclusivity and diversity so that under-represented sectors
eye on problems that have a more direct monetary payoff, of the population have unimpeded and equal access to
such as quantum chemistry. While many of the problems QIS education and training. It is also crucial to ensure
between different science areas are quite similar, the HEP retention of talent such that a well-trained new genera-
community might have unique needs that might not be tion of scientists in quantum simulation for HEP have
fully addressed by the private sector. This would be similar attractive permanent positions to look forward to, since
to the situation in the lattice-QCD program, where building otherwise they will leave the field for lucrative positions in
dedicated hardware proved to be important to fully realize the private sector. This can come through faculty positions
the science goals, as well as contributed to a development at universities, permanent scientist positions at national
and flourishing of parallel architectures. A careful study laboratories, and ideally joint positions between the two,
over the next decade will be needed to assess the hard- following a successful model in joint faculty and staff
ware needs of the HEP community, which will inform how positions in nuclear physics over the past decades.
many HEP resources will need to be allocated for hardware Finally, quantum computational physicists will be
development. essential in developing application-specific algorithms and
Finally but crucially, advancing quantum simulations software. Academic institutions would need to recognize
of processes essential to HEP research objectives requires the value they bring to the physics community in general,
a diverse and inclusive quantum ready workforce with and the HEP field in particular, and not treat them as sole
skills that extend significantly beyond those traditionally computing practitioners, rather as physicists who deserve
in HEP, making workforce development a key compo- (full or joint) permanent positions in physics. Along the
nent of the program. The skills required for quantum same lines, R&D will continue to be a major part of quan-
simulations of HEP processes [316] include HEP phe- tum computing research in HEP, and physics results with
nomenology, quantum field theory and quantum mechan- impact on the HEP theoretical and experimental programs
ics, lattice field theory, HPC, statistical analysis, exper- may not materialize within the career span of a graduate
imental design and optimization, machine learning and student, a post-doc, or even a junior faculty. The value of
artificial intelligence, software-stack development, quan- this endeavor may be established in the long run, hence it is
tum and classical-computing algorithms, quantum circuit important to recognize the nature of this research, and not
design, implementation and optimization, and more. Inte- penalize (junior) scientists that invest in developing novel
gration with quantum hardware development through HEP frameworks that can enable future physics achievements.
co-design efforts will further broaden this skill set. Devel-
oping a skilled workforce has to be achieved through the
ACKNOWLEDGEMENTS AND FUNDING
natural realignment of some of the existing HEP work-
INFORMATION
force, but also through the recruitment and training of
junior scientists. The intrinsically interdisciplinary nature We are grateful to the members of the community
of quantum simulation for HEP requires collaborating with who endorsed this document as named in Appendix O,
expertise in QIS, computer science, applied mathematics, as well as to Mohsen Bagherimehrab, Aniruddha Bapat,
statistics, material science, nuclear physics, and more, with Shailesh Chandrasekharan, Lena Funcke, Jad Halimeh,
scientists, engineers, and developers residing in national Aram Harrow, Philipp Hauke, Joshua Isaacson, Karl
laboratories, universities, and technology companies. His- Jansen, Natalie Klco, Michael Kreshchuk, Andreas Kro-
torically, HEP scientists have contributed to, or moved nfeld, Norbert Linke, Vincent Pascuzzi, Indrakshi Ray-
into, other areas of research, such as computing, “big data,” chowdhury, Enrique Rico Ortega, Ananda Roy, Federica
and device fabrication, and have shaped the development Surace, Wei Xue, Erez Zohar, and Martin Zwierlein for
of those areas in substantial ways. The same is anticipated valuable feedback on an earlier draft of this Roadmap.
in the emerging quantum era. Christian Bauer is supported by the U.S. Department
Education and training programs for the skills men- of Energy’s (DOE’s) Office of Science under contract
tioned above will have to be distributed between univer- DE-AC02-05CH11231. In particular, support comes from
sities, national laboratories, and the private sector. The Quantum Information Science Enabled Discovery (Quan-
educational and training pipelines need to be both strength- tISED) for High Energy Physics (KA2401032).
ened and expanded in the area of QIS, and coordina- Zohreh Davoudi is supported in part by the U.S.
tion between the different sectors is required, since all DOE’s Office of Science Early Career Award, under
have an important role to play. Given the scope of the award no. DE-SC0020271, the DOE’s Office of Science,

027001-14
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

Office of Advanced Scientific Computing Research, Quan- Yannick Meurice is supported in part by the U.S. DOE’s
tum Computing Application Teams program, under field- Office of Science, Office of High Energy Physics Quan-
work proposal number ERKJ347, and the Accelerated tISED program, under award no. DE-SC0019139.
Research in Quantum Computing program under award Christopher Monroe is supported by the NSF’s STAQ
DE-SC0020312. She also acknowledges support from program, under award PHY-1818914 and the DOE’s Office
National Science Foundation Quantum Leap Challenge of Science, Office of High Energy Physics, under Award
Institute for Robust Quantum Simulation under Grant No. No. DESC0019380.
OMA-2120757. Guido Pagano acknowledges support by the DOE’s
A. Baha Balantekin is supported in part by the U.S. Office of Science, Office of Nuclear Physics, under Award
DOE’s Office of Science, Office of High Energy Physics, No. DE-SC0021143. He is further supported by the NSF
under Award No. DE-SC0019465. CAREER Award (Award No. PHY-2144910), the Army
Tanmoy Bhattacharya is partly supported by the Los Research Office (W911NF21P0003), and the Office of
Alamos National Laboratory and the U.S. DOE’, Office Naval Research (N00014-20-1-2695, N00014-22-1-2282).
of Science, Office of High Energy Physics, under Con- John Preskill is supported in part by the U.S. Depart-
tract with Triad National Security, LLC, Contract Grant ment of Energy, Office of Science, Office of Advanced
No. 89233218CNA000001 to Los Alamos National Labo- Scientific Computing Research, (DE-NA0003525, DE-
ratory. SC0020290), and Office of High Energy Physics under
Marcela Carena, Henry Lamm, and Ying-Ying Li are Awards DE-ACO2-07CH11359 and DE-SC0018407. He
supported by the DOE through the Fermilab QuantiSED also acknowledges funding provided by the Institute for
program in the area of “Intersections of QIS and Theo- Quantum Information and Matter, an NSF Physics Fron-
retical Particle Physics.” Fermilab is operated by Fermi tiers Center under NSF Grant No. PHY-1733907, the
Research Alliance, LLC under Contract No. DE-AC02- Simons Foundation It from Qubit Collaboration, and the
07CH11359 with the United States Department of Energy. Air Force Office of Scientific Research under Grant No.
Henry Lamm, David Van Zanten, and Silvia Zorzetti FA9550-19-1-0360.
are supported by the U.S. DOE, Office of Science, The work of Enrico Rinaldi is partly supported
National Quantum Information Science Research Centers, by the Royal Society International Exchanges Award
Superconducting Quantum Materials and Systems Center IEC/R3/213026. He is further supported by Nippon Tele-
(SQMS) under the Contract No. DE-AC02-07CH11359. graph and Telephone Corporation (NTT) Research.
Ying-Ying Li is further supported by the National Science Martin Savage is supported in part by the U.S. DOE’s
Foundation of China through Grant No. 12047502. Office of Science, Office of Nuclear Physics, InQuba-
Wibe A. de Jong was supported by the DOE’s Office tor for Quantum Simulation (IQuS) under Award No.
of Science, Office of Advanced Scientific Computing DE-SC0020970.
Research Accelerated Research for Quantum Computing George Siopsis acknowledges support by the Army
Program under Contract No. DE-AC02-05CH11231. Research Office under Award W911NF-19-1-0397, the
Patrick Draper and Aida El-Khadra acknowledge sup- National Science Foundation under Award DMS-2012609,
port from the DOE’s Office of Science QuantISED pro- and by the Defense Advanced Research Projects Agency
gram under an award for the Fermilab Theory Consortium (DARPA) Optimization with Noisy Intermediate-Scale
“Intersections of QIS and Theoretical Particle Physics.” Quantum devices (ONISQ) program under Award No.
Aida El-Khadra is further supported in part by the Simons W911NF-20-2-0051.
Foundation under their Simons Fellows in Theoretical Kübra Yeter-Aydeniz was supported by MITRE Corpo-
Physics program. ration TechHire Program, approved for public release with
The work of Masanori Hanada is partly supported Case No. 21-03848-2.
by the Royal Society International Exchanges award
IEC/R3/213026.
The work of Dmitri Kharzeev is supported in part
STATEMENT OF CONFLICT OF INTEREST
by the U.S. DOE’s Office of Science Grants No. DE-
FG88ER40388 and No. DE-SC0012704, and Office of A number of the authors of this Roadmap have a finan-
Science, National Quantum Information Science Research cial interest in the field of quantum computing and quan-
Centers, Co-design Center for Quantum Advantage under tum simulation: Nate Gemelke is the Chief Technology
Contract DE-SC0012704. Officer of QuEra Computing Inc., Junyu Liu is a scien-
Junyu Liu is supported in part by International Business tific advisor for qBraid Corporation, Mikhail Lukin is the
Machines (IBM) Quantum through the Chicago Quan- co-founder of QuEra Computing Inc., Christopher Mon-
tum Exchange, and the Pritzker School of Molecular roe is co-founder and chief scientist at IonQ Inc., John
Engineering at the University of Chicago through AFOSR Preskill is an Amazon Scholar affiliated with the Amazon
MURI (FA9550-21-1-0209). Web Services Center for Quantum Computing, and Kübra

027001-15
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

Yeter-Aydeniz is the Lead Quantum Algorithms Specialist short-distance process is computed to a given order in
at the MITRE Corporation. perturbation theory. Much work has been devoted to calcu-
late short-distance processes as precisely as possible, and
many processes are available at the second order in per-
APPENDIX A: PHYSICS DRIVE: COLLIDER
turbation theory, and some even at the third order. For a
PHENOMENOLOGY
recent review, see Ref. [9] and references therein. Care
Probing the Standard Model at the highest possible needs to also be taken to avoid double counting when
energies is usually achieved using high-energy particle combining parton-shower algorithms with short-distance
colliders. In such experiments, two initial-state particles calculations at higher orders in perturbation theory. While
(typically electrons, positrons, protons, or antiprotons) are the combination of these ingredients allows simulation of
accelerated to very high energies and then collide with fully exclusive scattering processes, the presence of the
one another. In this collision, the initial-state particles can various approximations made imply that the resulting dis-
scatter inelastically, and the large kinetic energy in the col- tributions can only be trusted for certain observables, and
liding particles can be used to create intermediate, massive it is difficult to estimate the uncertainty in the obtained
particles. By studying the decay products of these colli- predictions.
sions, one can infer what kind of intermediate particles In order to reduce the sensitivity to the details of the
were produced. A main use of particle colliders is to look modeling of hadronization effects and to the approxi-
for deviations from the SM predictions, which are expected mations made in the parton shower, comparisons of the
at some scale due to the inability of the SM to describe obtained simulations to experimental measurements are
well-established facts about nature, such as the existence limited to observables that are “sufficiently inclusive.”
of dark matter and the matter-antimatter asymmetry. While parton showers allow one to calculate less inclu-
Particle collisions are notoriously difficult to describe. sive observables, the results are much more dependent on
On one hand, this is due to the fact that particle collisions the particular choices made in the modeling, and therefore,
are governed by physics at widely different length scales. will have significant uncertainties.
While particle collisions are often aimed at discovering Quantum computers hold the promise to simulate scat-
processes happening at the highest energies (shortest dis- tering processes from first principles, and in principle,
tances), the observed distribution of final states is affected without any uncontrolled approximations. The basic idea,
by physics at energies ranging from the large kinetic which is explained in more detail in Appendix F, is to
energy in the colliding partons (quark and gluon con- discretize the continuous spatial dimensions and to digi-
stituents) all the way to low energies describing the binding tize the continuous field values [114,317,318]. This turns
of partons into hadrons. On the other hand, high-energy the infinite-dimensional Hilbert space of the field theory
collisions typically give rise to a large number of final-state describing the SM into a finite-dimensional one, which
partons (even before hadronization and decay), making can be described by the rules of regular quantum mechan-
them too complicated to be calculable using perturbative ics. Since the Hilbert space is exponential in the number
techniques based on, e.g., Feynman diagrams. of lattice points required, it is far too large to allow such
For these reasons, particle collisions are nowadays a computation using classical computers. However, it is
described theoretically using various types of approxima- believed that all relevant ingredients required to compute
tions, each valid for a certain energy range in the process. the S matrix can be calculated on a quantum computer
Processes at the highest energies typically involve only a using resources that scale only polynomially with the num-
small set of final-state particles, allowing a perturbative ber of lattice sites [13]. This was demonstrated explicitly
evaluation of the full quantum mechanical amplitudes. The for a scalar field theory [14], but is believed to also remain
production of the large number of additional partons is tra- true for more complicated field theories, such as gauge
ditionally described by a parton-shower algorithm [10–12], theories of the SM.
which is based on classical emission probabilities rooted As an alternative to the simulation of the full S
in a collinear approximation in the limit of infinite number matrix using quantum algorithms, one can also attempt
of colors (NC → ∞). Finally, one uses phenomenological to devise quantum algorithms for the parton shower and
models to describe how the resulting partons hadronize to the hadronization process. The idea of a quantum par-
form color-neutral hadrons. The parton-shower algorithm ton shower is to still work in the collinear approximation
is typically combined with some hadronization model to that underlies a classical parton shower, but to include
allow so-called exclusive event generators, which take a quantum interference effects that are not possible in a
partonic state produced in a short-distance process and traditional approach using classical probability distribu-
turns it into a fully exclusive final state containing only tions. For example, a quantum parton shower was shown
stable or long-lived particles, which can be observed to be able to include quantum interference effects arising
in a particle detector. For a review and discussion on from amplitudes with the same final-state particles, but
event generation in particle physics, see Refs. [7,8]. The different intermediate-particle flavors [225]. One can also

027001-16
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

hope to go beyond the Nc → ∞ limit of regular parton in effective theories with less resources compared to a
shower (with Nc being the number of colors), including the quantum simulation involving the entire collision. Exam-
quantum interference between different color structures. ples of such observables are diffusion coefficients, exclu-
Another approach that might be suitable for collider sive decay rates [319], parton distribution functions [134,
physics is to simulate the physics at a particular energy 216–218,320,321], hadronic tensor [216], transport coef-
scale on a quantum computer, while maintaining more ficient [322], and jet functions. These observables require
traditional approaches for physics at other scales. For substantially fewer quantum resources because they avoid
example, one could try to develop quantum algorithms manipulating late-time asymptotic states, and therefore,
for hadronization processes, which might allow one to represent interesting targets for simulation in the com-
go beyond the relatively simple models used in tradi- ing decade. Current estimates suggest that evaluation of
tional approaches. However, in order for such an approach the low-order transport coefficients for use in relativistic
to be meaningful, a proper separation of the various hydrodynamics require the smallest quantum resources,
energy scales present in a collider process is required. but the qubit and gate estimates are still substantial for
This is typically achieved using EFTs, as discussed QCD [164,322]. Obtaining such results for phenomeno-
in Appendix G. logical models such as the 2 + 1-dimensional quantum
Ising model will be a physically interesting near-term
target of quantum computation. Other interesting ques-
APPENDIX B: PHYSICS DRIVE: MATTER IN AND
tions in the development of quantum algorithms are the
OUT OF EQUILIBRIUM
effect of thermal fluctuations in hydrodynamics and finite-
Many open problems in HEP deal with the behavior volume effects in the quantum simulation of transport coef-
of strongly coupled fermionic matter at high density or ficients. With even fewer quantum resources, the dynamics
far from equilibrium. In the coming decade, results from of confinement and string breaking can be investigated
heavy-ion and proton collisions at the RHIC and the LHC in low-dimensional models [149,225,323] to improve
as well as gravitational-wave searches from the LIGO phenomenological models of parton showers [324] and
will lead to an unprecedented level of experimental data hadronization, e.g., the Lund model [325].
probing strongly interacting matter. Because of the strong Beyond computing nonperturbative inputs, the ability
coupling, one is typically forced to rely on the numerical of quantum simulations to perform real-time evolution
method of lattice QCD for first-principles predictions of will provide insight into the approach to equilibrium in
the state of matter. However, both in finite-density sys- strongly interacting systems [22]. There is tantalizing evi-
tems and for real-time dynamics, the use of the familiar dence from heavy-ion and proton collision experiments
Monte Carlo approach to lattice gauge theories is ham- at the RHIC and LHC that the strongly interacting mat-
pered by sign problems [5,6]. The quantum simulation ter thermalizes in a remarkably short period of time, a
of strongly interacting matter in and out of equilibrium fraction of a Fermi. This runs counter to naive scaling
holds the promise of avoiding such problems and can be arguments and suggests that our current understanding
performed efficiently, i.e., with resources that scale only of the dynamics of strongly interacting systems in their
polynomially in the size of the system by encoding the approach to equilibrium is missing fundamental insights.
entanglement of quantum states [19,20]. In finite-density One possible solution to this puzzle is that the dynamics
systems, such simulations should lead to precision theoret- of quantum entanglement plays a role in the equilibra-
ical results for the QCD equation of state and the behavior tion process [23–26]. Indeed, it has been proposed some
of phase transitions in strongly interacting matter [21]. time ago that a quench in an entangled system can lead to
Beyond their intrinsic value for better understanding QCD, apparently thermal behavior if only a part of the system is
the current theoretical uncertainty on these predictions are observed [326,327]. Studying such behavior in large, far-
anticipated to be the limiting factor in precision physics from-equilibrium strongly interacting quantum matter is
and searches for new physics in the coming decade. prohibitively difficult with classical computers. Therefore,
One area of real-time dynamics accessible to quan- the use of quantum computing represents a necessary tool
tum computers is the direct simulation of the collision in order to study the real-time evolution of entanglement
of leptons, hadrons, and nuclei, thus “solving” hadroniza- in strongly interacting matter.
tion in the sense of providing quantitative, testable pre- The insights from such research would enhance our
dictions, as discussed in Appendix A. Current resource understanding of the role of quantum information in HEP.
estimates suggest such computations require millions of In particular, it is known that entanglement spectrum
logical qubits and thus represent targets on a longer time [27], that is the spectrum of eigenvalues of the (nega-
scale. Instead, one can study these collision experiments tive of the logarithm of) reduced density matrix, contains
with phenomenological models composed of effective the- more complete information about the system than the
ories. A quantum computer provides a nonperturbative entanglement entropy [328,329]. The distribution of level
calculation of the low-energy or long-distance observables spacing in the entanglement spectrum can reveal whether

027001-17
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

and how the system thermalizes, and the evolution of Thanks to the “infinite” range of the neutrino-neutrino
entanglement-spectral gaps can signal phase transitions interaction in momentum space, this approximation can be
[28–31]. Furthermore, elastic and inelastic processes are rigorously justified when computing equilibrium proper-
shown to represent different rates of entanglement spread- ties at low energies (see, e.g., Ref. [345]) but its validity in
ing in the final state [54], as do the evolution of confined a more general out-of-equilibrium setting is yet not com-
versus deconfined phases of matter after a quench [330]. pletely understood. Nevertheless, the mean-field approach
While it is only recently that thermalization and associ- has been extensively utilized, revealing a large variety
ated questions in gauge theories have started to be explored of interesting physics such as synchronization, splits in
[31,331–334] using entanglement measures in classical the neutrino energy spectra, and early stage fast flavor
and quantum Hamiltonian-simulation studies, it is conceiv- oscillations. An important direction of current research
able that such explorations will gain a considerable boost is to clarify the range of validity of this approximation
as large-scale programmable simulators become available and to understand whether there exist situations where
[335–338]. Quench experiments may reveal interesting neutrino-neutrino correlation could change the mean-field
out-of-equilibrium features of the physical system [339, predictions in a qualitative way.
340] while being relatively straightforward to set up, as QIS tools could be utilized in the study of neutrino
demonstrated in quantum simulation experiments of spin astrophysics in two ways: one is using quantum comput-
systems that start to push the limits of classical computing ing to simulate the propagation of neutrinos, in particu-
[341,342]. Nonetheless, entanglement spectroscopy could lar, collective neutrino oscillations. Indeed early attempts
be relatively costly and recent ideas in shadow tomography with available digital quantum computers are encourag-
and related protocols [343,344] may provide economical ing [346,347]. However, a full digital simulation of the
ways to learn about entanglement structure of the final quantum many-neutrino systems is still in the far future.
state, see, e.g., Refs. [335,338]. Due to the simple structure of the neutrino-neutrino Hamil-
tonian, it is also possible, at least for simple geometries
and energy distributions, to map a neutrino system directly
APPENDIX C: PHYSICS DRIVE: NEUTRINO
into the degrees of freedom and interactions of quan-
(ASTRO)PHYSICS
tum simulators. Trapped-ion quantum devices are an ideal
In certain astrophysical environments, such as core- candidate for these near-term explorations due to the pos-
collapse supernovae and neutron-star mergers, the very sibility of exactly mapping the neutrino interactions (see,
large number of neutrinos present require taking neutrino- e.g., Refs. [35,36]). Further development of quantum simu-
neutrino interactions, both within the Standard Model lators to use qudits as fundamental degrees of freedom will
and beyond, into account. Proper description of neutrino also open the possibility of extending these studies beyond
propagation in such environments includes the effects of the simple two-flavor approximation.
neutrino mixing, forward scattering from the background In the near term, a more immediate application of QIS
particles, i.e., the Mikheyev-Smirnov-Wolfenstein (MSW) is to use tools, such as entanglement measures to assess
effect, forward scattering of neutrinos off each other and the limits of applicability of the mean-field approxima-
other collisions of neutrinos. The first three processes can tion. Since in the mean-field approximation, the entropy
give rise to collective neutrino oscillations, which have of entanglement for each neutrino vanishes, obtaining
been shown to potentially have an important impact in nonzero values of entanglement entropy in exactly solv-
supernova environment, both in the neutrino-driven explo- able simplified models would suggest how the mean-field
sion mechanism and in the ensuing nucleosynthesis in the approximation may be improved [348]. Indeed recent work
ejected material [32–34]. shows that away from spectral split energies, mean field
Correlations caused by neutrino-neutrino scattering pro- may be a good description, suggesting a hybrid approach
cesses result in a full many-body problem and an exact where many-body effects are explored for neutrinos with
solution to the dynamical evolution of flavor from a gen- energies around the spectral split energies [349]. It is
eral initial configuration will, therefore, require a com- also important to understand what observables would be
putational cost, which is exponential in the number of affected by quantum correlations, since the presence of
neutrinos involved. A peculiar property of the neutrino- entanglement in the many-body neutrino state does not
neutrino interactions, arising from the pointlike nature of necessarily imply an error in simple observables like
weak interactions in coordinate space, is that they are the individual flavor polarizations [350]. A better under-
extremely long ranged when represented in momentum standing of the evolution of entanglement with the size
space, resulting in large numbers of neutrinos’ momentum of the neutrino system being simulated is also critical
modes to be coupled together. to understand the crossover between the semiclassical
A simple approach would be to use a mean-field approx- regime, which can be explored using, for example, tensor-
imation in which a test neutrino interacts with a “mean network methods capable of describing weakly entangled
field” representing the influence of all the other neutrinos. many-neutrino states and a fully quantum regime

027001-18
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

requiring full-scale quantum simulations (see, e.g., include quantum effects. Multiple phenomena across cos-
Refs. [351–353]). mology and the early universe, including inflation, baryon
The measurement of fundamental properties of neutri- asymmetry, phase transitions, and dark matter need to
nos such as their absolute masses, mixing angles, and be described by the nonequilibrium dynamics of nonper-
the presence of CP violation is also a major goal of turbative quantum fields. At their core, these problems
the HEP community both in the U.S. and worldwide, demand tools capable of nonperturbatively time evolv-
and large-scale experiments, such as the Deep Under- ing the quantum fields, which is a challenging task. Alas,
ground Neutrino Experiment, have been commissioned to the only first-principles and systematic method to date,
achieve this goal. Besides its importance at a fundamental that is lattice field theory, is impeded in this endeavor
level, an accurate determination of these parameters would by seemingly intractable sign problems. To sidestep this
also inform simulation of neutrino interactions in astro- obstacle, state-of-the-art calculations assume adiabatic or
physical settings. In order to extract neutrino properties, near-equilibrium evolution, and/or perturbative field the-
long-baseline experiments rely on accurate determination ory. Such approximations may receive large corrections
of neutrino cross sections with the material used in their from far-from-equilibrium or nonperturbative effects that
detector. For the DUNE experiment, as well as the cur- are difficult to quantify. Obtaining these corrections or sim-
rently operating ones like MicroBoone, this is the 40 Ar ulating ab initio such phenomena constitute opportunities
nucleus. A full description of the neutrino-argon scat- for practical quantum advantage. Alongside the develop-
tering cross section is a formidable theoretical challenge ment of quantum hardware, theoretical developments are
due to the wide energy range required for the analysis of required in connecting classical- or perturbative-physics
the experiment, the necessity to have information about intuition with well-defined renormalized matrix elements
semiexclusive processes (such as neutron emission), as and algorithms to compute them. It is anticipated that
well as the open-shell structure of the 40 Ar nucleus, which addressing the problems discussed below with quantum
makes it challenging to describe accurately the properties hardware will, in the long run, change our understanding
of the target. of the early universe.
Classical techniques, including quantum Monte Carlo, In the inflationary paradigm, the universe experienced
coupled cluster, and various schemes based on Green’s a period of accelerated expansion brought on by quantum
function are currently being extended to address in part fluctuations, evolving as nearly classical fields, before ter-
this challenge (see Ref. [354] for a recent review of minating by nonperturbatively transferring energy to parti-
these theoretical efforts). Future quantum simulations can cle degrees of freedom through reheating and preheating.
help these efforts by allowing for both more efficient rep- These far-from-equilibrium and nonperturbative processes
resentations of the nuclear target’s wave function, and could leave imprints on the sky today [43–45]. While in
the extraction of both inclusive and semiexclusive reac- the long term, large-scale quantum simulations of quantum
tion cross sections with controllable uncertainties [37,38]. inflationary fields are desired [46–49], near-term studies
Early attempts with current digital quantum devices are could improve calculations of the quantum back reactions
encouraging (see, e.g., Refs. [38–42]) but to achieve [50] onto classical inflation fields and scalar-tensor per-
the required accuracy, large-scale error-corrected quantum turbations directly from nonperturbative quantum effects.
simulators are likely required. In the near term, it will Further near-term opportunities exist in using analog quan-
then be important, together with improving the scalabil- tum devices to simulate the dynamics of reheating with
ity of accurate algorithms, to understand in more detail ultracold Bose gases [46].
what type of semiexclusive data would be available by Generating the observed baryon asymmetry requires
quantum simulations and how to integrate this informa- nonequilibrium dynamics [51]. Potential sources of this
tion in the event-generator codes employed to analyze the nonequilibrium behavior include heavy-particle decays
experiment (see, e.g., Refs. [8,355] for recent work in [357], the Affleck-Dine mechanism [358], and first-order
this direction for classical simulations). Another impor- phase transitions [359,360]. Standard treatments rely upon
tant aspect that merits consideration in the future is how to the dynamics when the interaction rate is close to the Hub-
properly account for relativistic corrections, which become ble rate, which is essentially assuming near-equilibrium
important at large energy and momentum transfer (see, behavior [66]. Condensed-matter studies suggest that such
e.g., Ref. [356]). These directions will likely form the basis simplistic comparisons can be insufficient to ensure adia-
of an active research in quantum simulations of relevance batic evolution [67]. The state-of-the-art perturbative cal-
to the neutrino program. culations of the effective potential at finite temperature
may also suffer from non-negligible higher-order correc-
APPENDIX D: PHYSICS DRIVE: COSMOLOGY tions [361]. For a first-order phase transition, interesting
AND EARLY UNIVERSE phenomena such as particle production and gravitational-
The universe is inherently quantum, therefore wave generation are yet to be fully understood. Serving
predictions about earlier epochs of the universe should as inputs to the time evolution of classical fields, the

027001-19
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

properties of nucleating bubbles are usually extracted from of nature. We hope to resolve long-standing puzzles about
the effective potential. As recently stressed in the liter- what happens inside black holes, and about how black
ature [68], the phase-transition history defined solely by holes process information. We desire general principles
the conventional critical-temperature calculation can be that can explain the initial conditions in the history of the
misleading, and it is necessary to perform more detailed universe. We anticipate that progress in quantum gravity
calculations of the nucleation mechanism. Early quan- will teach us broader lessons applicable to other areas of
tum research should focus on rigorously defining matrix physics, and finally, studying quantum gravity is a fun
elements for bubble properties and exploratory calcula- intellectual endeavor!
tions of them [52–54]. Additionally, the interplay between Quantum gravitational phenomena are so elusive that
phase transitions and curved spacetime can be investigated one is forced to develop the theory with limited guid-
with the NISQ-era simulators to manifest nonequilibrium ance from experiments. It takes hubris even to try, and
behavior in low-dimensional toy models [55–58]. Stud- yet, remarkable progress is being achieved, particularly
ies of bubble nucleation and the growth of entanglement in the past few years. Many valuable lessons flow from
with energy and with the number of collisions have been the discovery of AdS/CFT duality 24 years ago [71], aug-
carried out using nonintegrable Ising spin-chain systems mented 15 years ago by the realization that, in the context
[54]. Given the limited digital quantum resources, ana- of AdS/CFT, bulk geometry emerges from boundary entan-
log quantum simulations of similar processes [362] would glement [72]. In a sense, quantum entanglement is what
also be beneficial. Moving beyond the near-equilibrium holds space together.
phase transitions, the quantum devices will also allow for Progress has been fueled by melding insights from the
the study of dynamical phase transitions, which arise in holographic correspondence with ideas from QIS. It is
nonequilibrium statistical mechanics [59–61]. discovered that the dictionary mapping bulk to boundary
Another mystery of the early universe is what accounts physics can be viewed as the encoding map of a quantum
for around 85% of the matter observed only via its gravi- error-correcting code [73,74]. Among other consequences,
tational effects. Light dark matter such as axions typically this viewpoint has sharpened our understanding of why
require a more complex and nonequilibrium history to be exact global symmetries are disallowed in bulk quantum
produced. For example, the strong-CP θ information and gravity [75]. We have learned that black holes are the
the temperature-dependent axion mass [62] are required most efficient possible scramblers of quantum informa-
to predict the relic abundance from misalignment mech- tion [76,77], and have leveraged that insight to deepen our
anism. To extract the mass information, QCD free energy understanding of quantum chaos more broadly [78].
as a function of the CP-violating phase and temperature It is further seen that computational complexity of a
has to be calculated. However, due to the presence of boundary theory can be related to geometrical proper-
the CP-violating phase, classical calculations suffer from ties in the bulk [79–81], and that a surprisingly simple
a sign problem. Though the dilute instanton gas model quantum system can have a holographic dual, which is
(DIGM) at high temperature or the interacting instanton helpful for understanding the system’s behavior [82,83].
liquid model (IILM) around the QCD phase transition More recently, by extending the connection between
has been explored in detail [63,64], it is unclear to what geometry and entanglement to include quantum effects
extent the DIGM and IILM are valid and how to control in the bulk [84], it is learnt how to compute the so-
their uncertainties. Quantum computers could be used to called Page curve, which describes how quantum infor-
compute the QCD free energy at finite temperature with mation escapes as a black hole evaporates [85,86].
a finite θ term [363,364]. With real-time simulations of Remarkably, thanks to the discovery of replica worm-
the early universe, quantum computers can also simulate hole contributions to the Euclidean path integral [87,88],
the nonequilibrium histories of light dark matter as well semiclassical computations validate the unitarity of black-
as reduce the systematic uncertainties involved. Quantum hole evaporation, without invoking any explicit description
simulation may allow probing the dynamics of other types of the black hole’s microscopic degrees of freedom.
of dark-matter-like topological defects [69], or primordial Despite this encouraging progress, much is still miss-
black holes [70] that are difficult to analyze because of their ing from our current understanding of quantum gravity.
relation to the strong-field theory. While the Euclidean path integral seems to be a surpris-
ingly powerful tool, it is not known how to formulate
it precisely in a theory of gravity. The knowledge about
APPENDIX E: PHYSICS DRIVE:
the quantum code relating bulk and boundary degrees of
NONPERTURBATIVE QUANTUM GRAVITY
freedom is yet incomplete. We do not fully understand
Constructing a complete and convincing quantum the- why bulk physics is local on distance scales small com-
ory of gravity is a grand challenge facing fundamental pared to the AdS curvature scale, which is related to our
physics. There are compelling reasons to pursue this quest. unsatisfying grasp of how quantum gravity works in
We crave a deeper understanding of the fundamental laws asymptotically flat spacetime or in de Sitter space. It is not

027001-20
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

yet known what happens at the singularity inside a black for advancing the tools needed to simulate dynamics in
hole, or even how to describe the experience of someone conformal field theories.
who falls through the event horizon. The computational More tractable, but still ambitious, targets with good
complexity of the dictionary that maps the region deep gravitational duals would be Banks, Fischler, Shenker,
inside a black hole to the region outside cannot yet be Susskind (BFSS) and Berenstein, Maldacena, Nastase
characterized with confidence. Finally, there does not exist (BMN) matrix models [89–91]. The information of the
a systematic way to identify quantum systems that admit gravitational geometry is encoded in matrix degrees of
useful holographic dual descriptions. freedom [89,92,93]. For example, the low-energy states
Quantum simulations, in both the near term and the are expected to describe M -theory black holes or black
longer term, can help to fill these gaps in our current under- zero branes [90]. Although Monte Carlo simulations on
standing. Thanks to holographic duality, bulk gravitational classical computers provided nontrivial test of this con-
phenomena can be described in a completely different lan- jecture (see, e.g., Refs. [94,95] for state-of-the-art results),
guage that does not involve gravity at all. Instead, the the details of the encoding of geometry into matrices have
“dual” boundary quantum system consists of many par- been out of reach. Quantum simulation would provide
ticles strongly interacting with one another. In principle, a practical tool in this problem. For example, a quan-
this boundary system can be simulated using a quantum tum state describing both a black hole and a probe D0
computer, opening opportunities for exploring quantum brane can be obtained by constraining some matrix entries
gravity in laboratory experiments. Furthermore, duality [93], and the motion of such a probe should be described
is a two-way street: on the one hand, experiments with by the black-hole geometry created by other degrees of
quantum devices might illuminate properties of quantum freedom [71]. Classical machine learning and hybrid quan-
gravity that are analytically intractable. On the other hand, tum classical algorithms have been successfully applied
by interpreting the behavior of many strongly interacting to simple matrix models [365,366], and quantum simula-
particles in terms of gravitational phenomena, one might tions including quantum machine learning would enable
better understand and control that behavior. studies of the full BFSS and BMN models. Real-time evo-
For example, properties of emergent geometry and grav- lution associated with the motion of D0 branes in the
itational back reaction in the bulk can be accessed by black-hole geometry, or simpler processes such as the scat-
exploring the entanglement structure of the dual bound- tering of a small number of D0 branes [367–369], would
ary theory. A particular challenge is understanding why the provide valuable clues to understand quantum gravity.
bulk gravitational theory is (approximately) local, given Quantum entanglement between color degrees of freedom
that operators, which are spacelike separated in the bulk can also be studied, and it may lead to a generalization
correspond to operators in the boundary dual that act of Ryu-Takayanagi approach that is based on the split-
on overlapping regions. Bulk operator commutation rela- ting of spatial regions. In principle, such approaches that
tions can, in principle, be studied in the boundary theory focus on matrix degrees of freedom can also be applied to
via experiments that probe transport or linear response. four-dimensional (4D) supersymmetric Yang-Mills theory.
Quantum corrections to semiclassical gravity, especially Furthermore, various supersymmetric systems including
nonperturbative ones, are difficult to compute analytically. 4D supersymmetric Yang-Mills theory can be regularized
Eventually, it may be possible to measure such corrections using matrix models, hence the quantum simulation of
in studies of the dual theory. matrix models can be the first step toward quantum simu-
To realize this vision of investigating quantum grav- lating supersymmetric QFTs, see Ref. [370] and references
ity through laboratory experiments, three ingredients are therein.
needed: (i) A (nongravitational) quantum system that has a Meanwhile, studies of the Sachdev-Ye-Kitaev (SYK)
gravitational dual, at least approximately. (ii) A proposal model [82,83], describing many fermions with strong all-
for realizing the quantum system in a feasible experi- to-all couplings, have alerted us that more accessible mod-
ment. (iii) A dictionary relating gravitational phenomena els can have useful gravitational duals worthy of further
of interest to measurable observables in the quantum sys- investigation. This development invites us to contemplate
tem. The best understood case is conformally invariant table-top experiments in which quantum information resid-
four-dimensional maximally supersymmetry SU(N ) Yang- ing in a complex many-particle system is first scrambled
Mills theory, such that the bulk curvature radius is large and then refocused, a phenomenon best understood in the
compared to the string scale when the boundary theory is dual bulk picture as transmission of quantum information
strongly coupled, and bulk effects higher order in the grav- through a wormhole in space [371,372]. Thus, gravita-
itational constant GN are suppressed when N is large [71]. tional intuition may guide our interpretation of dynamics
Admittedly, simulating this boundary theory accurately in strongly coupled many-particle systems even in the rel-
will require a large-scale fault-tolerant quantum computer, atively near term, especially for systems with long-range
which might not be available for decades, but this even- couplings [373–375]. Realistically, the goal in the near
tual goal provides one strong incentive (among many) term should be to light the way toward progress in the

027001-21
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

more distant future. Though spectacular insights into quan- properties of nucleons and light nuclei [99,100,386,387]
tum gravity may not be expected in the next 10 years, to inform astrophysical models and high-energy collider
the community can develop tools, methods, and insights experiments, and in understanding the dilute QCD mat-
that will lay foundations for quantum technology, enabling ter at finite temperatures [101–103] to shed light on the
profound advances in fundamental physics. phases of strongly interacting matter. Beyond QCD, lattice
field theory has been used to explore BSM theories in the
APPENDIX F: UNDERLYING SIMULATIONS: nonperturbative regime [388].
SIMULATING QUANTUM FIELD THEORIES Beside the exponential growth of the Hilbert space of
QCD as a function of system’s size, the statistical nature
Quantum field theory is an elegant mathematical frame- of the lattice-QCD method means that only a finite sam-
work that combines quantum mechanics and special rel- ple of (infinite) quantum configurations are produced and
ativity. The development of gauge-field theories through processed to estimate expectation values. This approach
various stages of conceptual and mathematical progress, suffers a significant drawback: with finite statistics, if con-
along with abundant experimental verifications, marked tributions to the system’s partition function with oscillating
the birth of the Standard Model of particle physics in the signs arise, statistical averages cannot be estimated reli-
20th century. Given that nature has chosen gauge theo- ably, leading to an infamous sign problem. This means
ries as the mechanism governing subatomic particles up that lattice-QCD calculations are bound to be performed
to length scales probed by experiments to date, physi- in imaginary time so as to allow a sign-problem-free
cists continue to believe that such theories, in one form or sampling using an Euclidean action. Moreover, in finite-
another, are strong candidates for physics beyond the SM density systems with a fermionic chemical potential, the
[376,377], including that of the dark sector [378–381]. Fur- probability distribution used in the Monte Carlo sampling
thermore, effective descriptions developed to describe cer- of quantum configurations is oscillatory even in Euclidean
tain limits of the SM, i.e., effective field theories, are a form spacetime and introduces a sign problem. A closely related
of QFTs. Moreover, there exist intriguing connections problem is an exponential signal-to-noise degradation in
between quantum gravity and conformal field theories. nuclear correlation functions, challenging precision lattice-
QFTs are, therefore, the backbone of HEP, and any attempt QCD calculations of nuclei. This limits the accuracy and
at simulating nature from first-principles amounts to sim- the precision of theoretical predictions for a range of
ulating quantum fields and their interactions. Perturbative HEP experiments that use hadron and nuclei as target.
methods have proven powerful in accurate predictions of Such limitations also mean that many fundamental ques-
the behavior of subatomic particles, e.g., at particle col- tions regarding equilibrium and nonequilibrium phases of
liders. However, for strong interactions in the low-energy QCD, including the mechanism of thermalization, hydro-
regime, where features such as confinement and hadroniza- dynamization, fragmentation, and hadronization in hadron
tion arise, one needs to apply nonperturbative methods collisions and early universe will remain unexplored, as
as the interaction strength becomes sizable. Furthermore, are a range of other critical questions in HEP, as detailed
in systems where electroweak and strong interactions are in this Roadmap. With this understanding, it is essential
both in play, such as in hadrons and nuclei, consistent to seek alternative computational paradigms that approach
inclusion of both interactions is required when handling these problems fundamentally differently.
the strong interactions nonperturbatively A primary question is the following: can a lattice-
field-theory program based on quantum simulation be
1. Conventional lattice field theory and the case for fully developed to complement and expand the conven-
quantum simulation tional program? To answer this question, one must recall
A reliable nonperturbative tool to simulate QFTs is lat- the course of developments in lattice QCD over multi-
tice field theory [318,382–385], which when applied to the ple decades. It consisted of, first of all, formally defining
theory of strong force is called lattice quantum chromo- the QCD path integral and observables in a finite dis-
dynamics (QCD). Lattice QCD is a numerical technique crete spacetime in such a way that as many symmetries
that systematically estimates, via Monte Carlo sampling, as possible are kept, or systematically recovered, in the
the quantum mechanical correlation functions of hadrons, continuum infinite-volume limit, starting from the pioneer-
nuclei, and finite matter from interactions among con- ing work of Wilson [389]. It is now proceeding to connect
stituent quarks and gluons. It has led to some of the most Euclidean finite-volume quantities to Minkowski infinite-
impressive computations in theoretical physics, from the volume amplitudes in the few-hadron sector, starting from
determination of muon’s anomalous magnetic moment to the pioneering work of Lüscher [104,105], along with
studies of light and heavy meson decays for testing the SM many other theoretical advances. It also consisted of devis-
of particle physics and searching for violations of funda- ing algorithms that, over time, scaled better with system’s
mental symmetries in nature [96–98]. It has also led to parameters and took advantage not only of advances in
progress in illuminating spectral, structure, and reaction applied mathematics and computer science but importantly

027001-22
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

of physics input, such as expression of symmetries and There are a number of frameworks developed for
constraints, and renormalization group and scale separa- systematically converting infinite-dimensional field theo-
tion, to make seemingly impossible computations possible. ries to a finite-dimensional counterpart. As an example of
Furthermore, it relied on adjusting algorithms and com- the choices encountered, consider a scalar field theory dis-
pilations to the hardware architecture, and remarkably in cretized on a spatial lattice. One may proceed by digitizing
instances, impacted the development of computing archi- and bounding the field and its conjugate variable [14], or
tecture itself via a co-design process, see Appendix L. alternatively can quantize the theory in terms of harmonic-
A quantum-simulation-based lattice-gauge-theory pro- oscillator excitations and bound the allowed occupation of
gram, similarly, should require developments in all these the oscillator modes [390]. Additionally, for simulations
three areas: (i) theoretical foundation, (ii) algorithmic in the low particle-number sector, a single-particle digi-
research, and (iii) hardware awareness, motivating the case tization may prove more economical [49]. Each of these
for co-design of dedicated QFT simulators. approaches may result in different rates of convergence to
the predictions of the infinite-dimensional theory, as well
as different resource requirements in simulation. The situa-
tion for gauge theories is more involved given the presence
2. Theoretical developments for quantum simulation of local gauge symmetries and their expression in basis
of QFTs states, and a number of leading ideas are currently being
The most common framework to compute static and explored. These include the following.
dynamical observables on quantum hardware is the Hamil-
tonian framework. This is because the unitary time evolu- (a) Global and local irreducible-representation bases.
tion can be naturally implemented on a quantum device, The starting point of these approaches is the Hamil-
either in a continuous (analog) manner or a digital (gate- tonian formulation of lattice gauge theories by
based) manner. While Kogut and Susskind laid the ground Kogut and Susskind [114]. The time variable is con-
for a Hamiltonian formulation of lattice gauge theories tinuous and a partial gauge fixing is performed by
in the 1970s [114], for the sake of quantum simulation, choosing A0 = 0, where A0 is the temporal compo-
more considerations are in play. The infinite-dimensional nent of the gauge field. A0 is nondynamical in the
Hilbert space of gauge bosons must be truncated and one Kogut-Susskind Hamiltonian. Therefore, it appears
may wonder whether electric field, magnetic field, or some as a Lagrange multiplier for the Gauss-law opera-
dual representation will lead to faster convergence to the tor, which is then required to vanish when acting
exact theory toward the continuum limit. Furthermore, the on the physical states. Since the Gauss law is a
local Gauss laws must be imposed on the Hilbert space statement on the divergence of (color) electric field,
(or more generally, gauge invariance should be preserved the electric field or irreducible representation (irrep)
either directly upon the Hilbert space or dynamically), or basis comes with advantages in applying the Gauss
else the simulation may explore a vast unphysical Hilbert law. For example, by analytically solving the Gauss
space due to algorithmic or hardware imperfections. Last law at every lattice site, only Casimir eigenvalues
but not least, the continuum and infinite-volume limits of are left as dynamical variables, from which a new
observables, static or dynamical, in the Hamiltonian frame- Hamiltonian matrix can be formed that has a lower
work, must be understood. We briefly discuss each of these dimension and does not involve all or some of the
theoretical directions, and enumerate avenues for progress unphysical transitions. This process can be done
in the coming decade. globally, with a cost that scales exponentially with
The Hilbert space of local QFTs is infinite dimensional the size of the system and is impractical for sizable
since, ignoring mathematical subtleties, the states can be simulations. It can also be done locally or semilo-
described by functions from physical space into a target cally, in which case the classical preprocessing is
field space. The physical space is a continuum and of scalable with the system’s size, but unphysical tran-
infinite extent. For fermionic theories, the target space is sitions among various (semi)local blocks are still
finite dimensional locally, while for bosonic theories like plausible and must be eliminated in the simulation
gauge theories, the target space is also continuous. On the algorithm at a cost. Applications of this approach
other hand, quantum computers realized as discrete sys- to U(1), SU(2), and SU(3) pure gauge theories in
tems can simulate only finite-dimensional systems, and the context of quantum simulation have appeared in
this requires discretizing and digitizing and bounding both recent years [115–118].
physical space and the target space. The original field the- (b) Prepotential and loop-string-hadron formulations.
ory is then realized as a double limit of removing both The starting point of these formulations is still
of these discretizations and bounds. To estimate resources Kogut-Susskind Hamiltonian of non-Abelian LGTs,
required to simulate a theory to a desired precision, one and one works in the irrep basis, in which the elec-
needs to understand these limits. tric Hamiltonian is diagonal, while the gauge-matter

027001-23
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

coupling and magnetic Hamiltonians are nondiag- below O(107 ) and the lattice spacings a  0.06 fm
onal. Prepotential formulation amounts to breaking [396,397]. As quantum resources improve, system-
the representation of the gauge link operator to left atic improvements to the Hamiltonians are possible
and right Schwinger bosons of the SU(N) theory, [127,391,392,394].
and building SU(N)-invariant operators from these (d) Magnetic or dual representations. A formulation in
at each site [119–122]. By coupling prepotentials the magnetic basis, where the magnetic Hamiltonian
to fundamental fermions, one can construct gauge- is diagonal, requires much fewer basis states to be
invariant bosonic and fermionic operators, the so- retained in the truncation, yielding a much more effi-
called loops, strings, and hadrons [123]. The loop- cient representation at weak couplings relevant for
string-hadron formulation, therefore, expresses the the continuum limit. The dual nature of the electric
non-Abelian dynamics in terms of strictly charge- and magnetic basis [126] has allowed a magnetic
conserving underlying operators. The ordinary non- basis to be constructed for a compact Abelian gauge
Abelian Gauss-law constraints are made automatic, theory [127,399]. Starting from an electric basis,
though auxiliary U(1) constraints are introduced which keeps a large number of electric basis states,
and imposed to ensure the equality of the group and converting this to a magnetic basis using a
Casimir on the broken links. The loop-string-hadron Fourier transform, the resulting magnetic basis can
Hamiltonian is naturally expressed as a sum of one- then be truncated, giving a much better descrip-
sparse terms, which could benefit time-evolution tion at small couplings than an electric basis of the
subroutines [285]. The development of the loop- same dimension, while performing much worse at
string-hadron formulation for QCD is among the large couplings. Another basis for the same com-
immediate next goals of this program. pact U(1) theory exists [128], in which the magnetic
(c) Group-element basis and discrete subgroups. The and electric basis are related to each other by a
magnetic Hamiltonian on the lattice is defined with simple Fourier transform, and the scheme is shown
semilocal operators involving the product of links to work at both small and large couplings alike.
along a closed path such as a plaquette. These have The development of dual bases for non-Abelian
a nontrivial action on the states expressed in the gauge theories will be an important next step for
irrep basis. Furthermore, the magnetic Hamiltonian the field, but early efforts indicate that such dual
dominates the dynamics of U(1) and SU(N) gauge formulations often lead to more complex and gen-
theories toward the continuum limit, and choosing erally nonlocal electric Hamiltonians [129]. The
an irrep basis requires retaining a large number of benefits of such dual formulations, therefore, must
electric-field excitations in this limit, hence increas- be thoroughly examined in the context of quantum
ing the computing-resource requirement. One may, resources required to achieve given accuracy.
therefore, desire to work in the group-element basis (e) Tensor renormalization group. A complementary
[124], which simplifies the simulation of both the approach starts with the standard Lagrangian for-
gauge-matter coupling and the magnetic Hamiltoni- mulation used in lattice gauge theory and uses
ans. Furthermore, by formulating in this basis, one character expansions (for instance, Fourier series)
maintains a close relation to standard lattice-field- developed in the context of strong coupling expan-
theory methods, which simplifies analysis [391,392] sions to rewrite partition functions and average
and development of algorithms [125,207,393,394]. observables in terms of products of traced ten-
However, quantizing and truncating the group ele- sors. In most situations of interest, this provides
ments in the SU(N) LGT is not straightforward a discrete reformulation that can be exploited for
and may violate the group symmetry. One approach quantum computing and can be verified at small
around this is to approximate continuous gauge volume with conventional methods. The tensors
groups by their crystal-like subgroups. This crystal- are local objects that contain all the information
lization reduces qubit costs [189,391,395–398] with about the model and its symmetries. They can be
realistic estimates for SU(3) LGT being approx- seen as the building blocks of various types of
imately 10 qubits per gauge link, and is agnos- computations. When continuous field variables are
tic to the particular Hamiltonian chosen [394]. involved, there is an infinite number of charac-
This remnant gauge symmetry simplifies renormal- ters (Fourier modes), but it has been shown that
ization issues, in particular from gauge-symmetry truncations of tensor sums preserve global and
violation. By proper choices of Hamiltonians and local symmetries [400,401], see Ref. [132] for a
actions, it has been demonstrated that for U(1) and recent review. The reformulation of lattice gauge
SU(N), systematic errors from this approximation theories was initially developed in collaboration
should remain negligible for quantum simulations with condensed-matter researchers [130], extending
for the forseeable future where qubit counts remain the method of Ref. [131]. The original motivations

027001-24
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

include clean configuration-space coarse graining one introduces a fictitious space dimension so that
[402] and absence of sign problem in the pres- the infinite local Hilbert space is built up as a direct
ence of a nonzero chemical potential [403] and a product of fixed-size Hilbert space on each site in
nonzero θ term [404,405]. Transfer matrix meth- the new direction. The advantage of this method is
ods also connect to the Hamiltonian approach and that the Hamiltonian that needs to be simulated is
quantum simulation [406–409]. There has been a local in this extended space, so the quantum circuits
considerable effort to extend this approach to mod- that implement it are potentially simpler. Recent
els with fermions [410–413], scalars [414–417], and work shows how the O(3) model with a θ vac-
other models [418–420] in various dimensions. Fur- uum can be studied in this framework [430]. It has
thermore, other tensor methods that were developed also been shown that the extra dimension may be
earlier in condensed-matter physics led to the use unnecessary in some cases [138]: the low-energy
of quantum information tools in approaching QFT sector of Hamiltonians tuned to a quantum critical
problems [110,111]. Moreover, near-term quantum- point are described by the QFT. Intuitively, at such
simulation algorithms might benefit from a com- a critical point the continuum local Hilbert space
bined approach of variational algorithms and tensor describes the state of the lattice system over a region
networks. The power of tensor networks could be of size given by the correlation length of the sys-
utilized by splitting large quantum systems by small tem, and so can be infinite as this length diverges.
subsystems [421], and QFTs could be natural targets The trick, of course, is to find the correct quantum
[422]. critical point, and these may not exist for theo-
(f) Light-front quantization. Hamiltonian QFT need not ries of interest. Interestingly, however, properties
be formulated on fixed time slices. In the light-cone like asymptotic freedom can arise [139,140] in this
quantization approach, fields are quantized along approach, and, often, with smaller resources, one
the light cone x+ ≡ t + z [423]. This is shown to can get EFTs that can be completed in the ultraviolet
result in a smaller number of physical degrees of by continuum perturbation theory. A systematic way
freedom compared with the canonical equal-time to explore the qubit-regularization approach within
quantization. The reason is that the sum of occu- the quantum link framework was discussed recently
pancies in a Fock state is upper bounded in the in Ref. [431]. Future work needs to extend this
light-cone quantization, since there is no possibil- approach to gauge theories and fermionic theories.
ity for an infinite number of left- and right-moving (h) Matrix models. Dimensional reduction can be used
massive particles, which can give rise to a net finite to map gauge theories to quantum mechanical mod-
momentum. Such an approach, that is related to els, while preserving some of the interesting non-
the single-particle quantization scheme mentioned perturbative dynamics and structure of the parent
above, puts quantum simulation of QFTs on a simi- QFT. With their much smaller Hilbert spaces, these
lar footing with the quantum simulation of quantum models are interesting physics targets for near-term
chemistry [424]. Nonetheless, subtleties associated simulations of gauge theories, complementary to
with the zero mode of massless fields and with UV- approaches based on digitization or other trunca-
IR mixing during renormalization [425] complicate tions of lattice gauge theories with small numbers
the scheme and require careful treatment, see, e.g., of sites. For a simple example, the reduction of
Refs. [426–428] for related progress. A digitiza- 1+1D QED with massive charged matter on a small
tion amenable to use on quantum computers for spatial circle leads to a quantum mechanical rotor
the SU(3) LGT was constructed recently [133–135]. model, which realizes the same ’t Hooft anoma-
The light-cone formulation is well suited for cal- lies and tunneling processes as the parent Schwinger
culating the properties of relativistic bound states, model. These properties are associated with slow
while the applications to the scattering problem are dynamics that can be seen in analog simulations
emerging [429]. on a single Rydberg atom [432]. In the context
(g) Quantum link models and qubit regularization. of ordinary 4D Yang-Mills theories, reduction on
There exist approaches for simulating QFTs with- a small spatial torus maps the gauge theory to a
out the need for an infinite-dimensional local Hilbert matrix quantum mechanics model with calculable
space. Such approaches to QFTs are well known in Hamiltonian [141], and the low-lying spectrum of
the condensed-matter literature and were brought to the matrix model can accurately reproduce the spec-
particle physics through the quantum link approach trum of the gauge theory obtained in Euclidean
via the idea of D theory [136,137]. It is argued that lattice simulations [142]. Thus quantum simulations
almost all quantum field theories can be obtained of matrix models may provide interesting insights
in this approach by formulating a lattice field to nontrivial dynamics of 4D gauge theories on
theory with a finite-dimensional Hilbert space, when near-term hardware. Related matrix models are also

027001-25
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

of broader theoretical interest, arising for example second approach is an active protection of the symme-
in nonperturbative formulations of string theory [89] tries as the system is evolved in the simulator. Examples
and other models of quantum gravity [143,144], see include adding a penalty term to the Hamiltonian propor-
Appendix E. Such models could also serve as a nat- tional to the (square of) Gauss-law operator to suppress
ural target for developing quantum algorithms and the leakage to the unphysical Hilbert space [149–153],
simulations [370,433]. performing random rotations during evolution with the
Gauss-law operator to average out the symmetry violation
As global and local symmetries played a crucial role in [154,155], adding to the Hamiltonian the Gauss-law opera-
the establishment of the SM, it takes a special effort to tor with properly chosen coefficients to separate out differ-
reconsider these questions in the context of Hamiltonian ent Gauss-law sectors in the spectrum [156,157] or similar
formulation of discretized and digitized approximations of techniques [158,159], using classical noise proportional to
QFTs. In the irrep-basis formulation of lattice gauge theo- the Gauss-law operator to suppress gauge-symmetry vio-
ries, a cutoff on the number of irreps retained respects the lation via a Zeno effect [160], a similar quantum approach
gauge symmetry except at the cutoff. On the other hand, in which quantum control is used to dynamically decou-
arbitrary digitizations and truncations in the group-element ple unphysical sectors during the evolution [161], and in a
basis may respect only a subset of symmetries or none. In more gate-based setting, using controlled operations to dis-
tensor reformulations of lattice models [132], it has been allow unphysical transitions between basis states [118]. As
shown that truncations of tensor sums preserve global and a verification step, one could also use oracles in the quan-
local symmetries [400,401]. More generally, symmetries tum circuit to detect Gauss-law violations and discard the
usually define a continuum limit that has universal aspects. result [147,148].
Reaching this limit in the most efficient and economical More research is needed to clarify the importance of
way, rather than closeness to a specific lattice model with symmetry-protected simulations and whether they will be
lattice artefact, must be considered the ultimate goal. more resource efficient, in general, compared with non-
The questions of gauge invariance and noise-robust protected counterparts. For example, the measure of merit
implementations of the Gauss laws for Abelian [126, should be the closeness to the exact evolution, and if a
401,434,435] and non-Abelian [148,285] local symmetries nonprotected algorithm has a faster approach to the exact
have received considerable attention in recent years. Even limits, it should be taken as the method of the choice.
if the digitized and truncated formulation of a gauge theory Furthermore, it is not clear that suppressing errors that
provides a (nearly) gauge-invariant Hamiltonian, simu- are associated with transitions to the unphysical sectors
lating the system under that Hamiltonian may break the will reduce the total error, as demonstrated in Ref. [162].
symmetries. Due to the condition of the Gauss law that is Furthermore, the incoherent noise appears to be the dom-
a constraint on the Hilbert space, even if the simulation inant source of simulation error in the NISQ hardware,
is launched in the physical sector of the theory, imperfec- and symmetry-protection protocols need to be general-
tions in the simulation algorithm or in quantum hardware ized to address such errors too. Last but not least, it may
can drive the system out of the physical subspace. For be that the gauge-theory simulation are robust to small
example, Trotterized evolution in a digital simulation may gauge-violating errors in the simulator, as demonstrated for
introduce errors that do not respect the symmetries, as does several quantum link models (QLMs) in Refs. [145,146],
an inaccurate engineering of the dynamics in an analog so an active symmetry enforcement with a resource over-
simulation. Coupling to the environment likely involves head may not be necessary after all. As the field moves
gauge-symmetry violating terms too. toward selecting the best theoretical formulations of gauge
There are two currently known approaches to gauge- theories for quantum simulation, all these questions need
invariant simulations, assuming that the simulation starts to be thoroughly addressed.
in the gauge-invariant sector but may evolve to other sec- A major part of the development of conventional lattice-
tors due to hardware or algorithmic imperfections. One gauge theory over the past few decades has been to
is to adopt a formulation, which is fully or partially quantify and mitigate systematic errors, such as those
gauge invariant by construction. Examples include purely associated with discretization, finite volume, excited-state
fermionic formulation of gauge theories coupled to mat- effects on ground-state properties, and quark-mass inputs
ter in 1+1 D with certain boundary conditions, where a (if set away from physical values for computational expe-
gauge transformation and the application of the Gauss diency). EFTs played a major role in these efforts as they
laws fully constrains the gauge degrees of freedom, leav- allowed construction of improved actions and observables,
ing only fermions, which can now interact nonlocally and finding reasonable extrapolation forms for certain
[285,436,437]. Another example is the loop-string-hadron quantities. Furthermore, studying finite-volume effects
formulation of the SU(2) LGT in d+1 D [123], which offered a powerful methodology to access few-hadron
incorporates the non-Abelian Gauss laws by construction scattering amplitudes that otherwise would not have been
but leaves an Abelian constraint to be satisfied locally. The accessible with the lattice-QCD technique [104–109]. It

027001-26
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

is conceivable that such a program will continue to grow ancillary qubits and entangling gates are desired. In the
in the Hamiltonian-simulation era too. In fact, investiga- fault-tolerant era of quantum computing, the overhead is
tions of infinite-volume and continuum limits of certain associated with error correction. Fault-tolerant implemen-
QFTs have emerged in recent years [393,438]. Besides tation of non-Clifford gates such as the T gate is known
new systematic uncertainties encountered in digital quan- to be more resource intensive than Clifford gates for error-
tum simulation, such as digitization of the time variable, correcting codes such as surface codes. Therefore, it is the
and the truncation errors in bosonic theories due to limited T-gate count that needs to be minimized in the far term.
qubit resources, constraints such as space discretization The most popular simulation algorithms to date are
in lattice formulations, and the finite extents of time and product formulas, which are based on Trotter-Suzuki
space remain in effect. Strategies to quantify and extrapo- decomposition of the time-evolution operator [439].  For
late them away may be rather different in a Hamiltonian- example, for a Hamiltonian of the form H = l=1 H (l) ,
(l)
simulation setting, but reliance on EFTs and improvement where the different
 H do not
r commute with one another,
 −itH (l) /r
programs may prove useful in this context too. As a result, the operator l=1 e approximates e−iHt up to an
more theoretical research is needed to address the ques-
tion of systematic uncertainties and their quantification for error that scales as O(t2 /r) for positive integer r and real
quantum simulation of QFTs. parameter t > 0. Higher-order formulas can be constructed
to enable more accurate simulations but at the cost of
increasing the circuit depth [440,441]. As a result, if the
3. Algorithmic research for digital quantum Hamiltonian is a sum of local or semilocal terms, the sys-
computing and resource analysis tem’s evolution can be implemented in polynomial time
The digital approach to quantum simulation offers con- [266]. With no ancillary overhead and simpler implemen-
trolled ways to prepare, evolve, and measure the states of tation, the product formulas may remain the simulation
a quantum system. Importantly, digital algorithms can be algorithm of choice in the near term. There has been a great
generally analyzed rigorously and their asymptotic or exact deal of progress in developing other simulation algorithms
resource requirement can be bounded given a desired accu- such as Taylor series expansion and linear combination
racy. Furthermore, a range of error-correction and error- of unitaries [442,443], quantum signal processing [444],
mitigation techniques applies to digital simulations. It is qubitization and block encodings [445,446], singular-value
important to invest in designing, analyzing, and improv- transformation [447], off-diagonal Hamiltonian expansion
ing suitable quantum simulation algorithms for HEP, and [448] and hybrid algorithms [153], which generally per-
particularly for QFTs of interest. In the context of quan- form more optimally asymptotically, but often involve
tum simulation of physical models in general, and QFTs significant (if scaling polylogarithmically) ancillary qubits
in particular, this section reviews the basic elements of a and more complex circuit implementations. On the other
digital approach to simulation and recent advancements. hand, better analytical approaches [441], taking advan-
Remaining open questions in simulation algorithms, and tage of the system’s locality and conservation laws [165]
in understanding their resource scaling will be further or inputting information about the initial state [166–169],
discussed. and empirical analysis of the performance in select cases
In a digital simulation, the system’s evolution is bro- [162,170,171] have resulted in considerably tighter bounds
ken to simpler implementable unitaries, and the way such a on product-formula errors in recent years. Research in the
digitization is performed defines the simulation algorithm. quantum algorithm community continues to improve the
Implementable in this context means unitaries for which current simulation schemes and to devise new strategies.
a decomposition exists that is composed of a number of QFT simulations will be a prime application of optimized
elementary quantum gates that is polynomial in terms of algorithms given their significant resource requirement.
problem size and error tolerance. For the purpose of this In the context of QFTs, among the first thorough algo-
section, we focus on the qubits as the quantum informa- rithmic approaches to quantum simulation is the seminal
tion units and a common choice of universal set of single- work by Jordan, Lee, and Preskill [13,14,449], which sets
and two-qubit gates as units of quantum processing (more up an evaluation of the scattering S matrix in an interacting
general entangling gates can be used as elementary gates field theory. This work demonstrates three primary tasks in
as well). More general choices, e.g., higher-dimensional quantum simulation: (i) initial-state preparation amounting
qudits or customized gates, will be discussed later in to preparing scattering wave packets, (ii) time evolution
the context of analog and hybrid approaches to quantum involving an adiabatic approach of the system to a fully
simulation. The algorithm’s figure of merit depends on interacting theory and evolving back to isolated wavepack-
resources that need to be minimized. In the near-term ets, and finally (iii) final-state measurement amounting to
computing model, qubit resources are scarce and entan- identifying quantities that can be optimally measured and
gling gates are lower in fidelity than the single-qubit gates. processed to access information about the scattering ampli-
Therefore, computations that require the least number of tude, without the need for costly full state tomography.

027001-27
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

While this algorithms shows the true advantage of a quan- the SU(2) and SU(3) lattice gauge theories in the irrep
tum computer, that is to enable a direct evaluation of real- basis, when approximating the time-evolution operator
time cross sections in QFTs, its resource requirement will via Trotterization for a maximum error  for an arbi-
likely prohibit its implementation for even small systems trary state, the gate count scales as ∝ dt3/2 (L/a)3d/2  −1/2
for the foreseeable future. Research is in progress to devise [163,455], although logarithmic corrections may prove
and benchmark less resource-intensive approaches to the important [164]. Here, d denotes the dimensionality of
scattering problem in the near term, including proposals space,  is the gauge-field truncation in the irrep basis,
for variational approaches [206], obtaining phase shifts in L is the spatial extent of a cubic lattice, and a denotes
prototype spin models via time delay [450], or the use of the lattice spacing. For an accuracy goal  = 10−8 and
Lüscher’s method [104,105] in extracting low-energy few- a lattice with tens of sites along each spatial direction,
body scattering parameters from energy spectra of particles as proposed in Ref. [164], the simulation requires hun-
in a finite volume, as is done in the conventional lattice- dreds of thousand to hundreds of million qubits and of the
QCD program [106–109]. A complete resource analysis of order of 1050 T gates and more. Given the unaccounted-
scattering problems in gauge theories, including QCD, is for theoretical uncertainties, acceptable values of  could
still lacking and the problem is complicated by the absence be orders-of-magnitude larger. A tighter bound could be
of fully developed and efficient state-preparation algo- obtained by considering state-dependent errors [166,167].
rithms for a range of nontrivial states in gauge theories, As the community pursues better approaches to simulating
from the interacting vacuum to the scattering wave pack- gauge theories, questions such as suitable formulations and
ets of confined hadrons. Promising progress is reported in practical, and perhaps more customized, implementations
recent years on state preparation in scalar [14,283,451], given the simulating hardware must be addressed. Hybrid
fermionic [282,449], and gauge-field theories [191], and digital-analog algorithms may prove valuable in scaling
it is plausible that customary state-preparation techniques, down the cost, but more work is needed to understand their
such as adiabatic state preparation [363], projective cool- time complexity, as discussed in the next subsection.
ing [452,453], and tensor-network-inspired ansatzes [454], The algorithm and its performance is closely depen-
combined with new customized QFT algorithms can lead dent upon the encoding of the degrees of freedom, and
to further progress in the coming years. this encoding is related to the choice of formulation as
A fundamental element of any quantum simulation discussed in the previous subsection. This choice, in par-
algorithm for QFTs is the implementation of the time- ticular, impacts the error-bound analysis of the simulation
evolution operator, given the formulation and basis states algorithm by systematically accounting for the trunca-
chosen and the encoding adopted for the various degrees of tion errors in bosonic field theories. The first attempts at
freedom. While the development of efficient algorithms for numerically investigating the truncation errors in scalar-
scalar field theory and Abelian and non-Abelian gauge the- field theory and Abelian and non-Abelian gauge theories in
ories has led to valuable detailed analyses of qubit and gate both irrep and group-element basis demonstrate the expo-
requirements for various formulations of a range of mod- nentially fast convergence of low-energy observables to
els [14,117,118,163,164], it is not clear that the devised the exact values [118,285,390,395], but fails to reach this
algorithms have the most optimal scaling and are suit- exponential scaling until larger truncation cutoffs are used
able for near- and intermediate-term computations. The for high-energy and long-time-evolved quantities [285].
situation has strong parallels in the classical-computing There exists analytical approaches to understand this expo-
world, where many initial algorithms, while they generated nential convergence in certain problems [455–458], and a
interest and guided the developments, were replaced by first analysis of error bounds in product formulas consider-
increasingly faster and more resource-efficient algorithms. ing this exponential convergence in noninteracting scalar
It also became clear that asymptotic scalings were not field theory and the SU(2) LGT in the irrep basis has
necessarily the best guide to accurate resource estimates, appeared [455]. Deriving analytical bounds for evolution
and identifying prefactors and benchmarking algorithms under truncated Hamiltonians is generally difficult but is an
in pursuit of learning their empirical performance were important step toward more reliable error-bound analysis
essential in advancing computational sciences and their of algorithms given the formulation used for the QFT. Even
applications. A similar trend is expected in the realm of in qubit-regularized and QLM approaches to recovering
quantum simulation. For example, it is becoming clear the continuum QFTs, which do not require extrapolation
that simulating non-Abelian gauge theories in the irrep in the dimensionality of the on-site Hilbert space, the rate
basis suffers from costly compilation of noncommutative of convergence to such continuum limits in connection
algebra of the group, necessitating many rounds of accu- to resources required needs to be examined thoroughly in
rate synthesis of nontrivial functions as the system evolves future investigations.
[164]. Both near- and far-term circuit implementations Besides the issue of quantifying truncation errors and
of such dynamical phases are costly and introduce non- convergence rate toward the continuum limit, one needs
negligible overhead to the simulation. For example, for to analyze the best available encoding strategies for both

027001-28
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

bosonic and fermionic degree of freedom, and such a ques- [335,338,343,344], for QFTs, particularly in QCD-like
tion must be addressed in the context of the QFT formu- theories with confined and composite asymptotic final
lation used and the simulation algorithm adopted. Among states.
proposed bosonic encodings are the standard binary encod- From this overview of the field and the outstanding
ing with which an integer b is encoded in approximately problems, it is obvious that algorithmic research will con-
log2 b qubits, Gray encoding with the same qubit count tinue to constitute a major component of the field of quan-
as the binary case, and the unary encoding with b qubits tum simulating QFTs. Progress relies on well-equipped
[459]. There is even more diversity with the fermionic field theories that can combine theory inputs with algo-
encodings. While fermionic on-site Hilbert space is finite rithmic needs, along with close collaboration with quan-
dimensional, implementing the Fermi statistics can be tum algorithm experts and digital-hardware developers, as
costly, leading to maximally nonlocal interactions among discussed in more details in Appendix L.
the qubits through Jordan-Wigner fermion-to-spin trans-
formation [460], or more local but qubit-resource-intensive
encodings [461–469]. In fact, the construction of many 4. Analog and hybrid approaches to quantum
local mappings resembles the structure of lattice gauge the- simulation of QFTs
ories, in which the interactions are locally mediated by In an analog approach to quantum simulation, the
the ancillary degrees of freedom but local “Gauss-law”- Hamiltonian of a theoretical model is mapped onto the
like constraints must be imposed on the Hilbert space Hamiltonian of an actual physical system engineered in
of the main and ancillary qubits, requiring robust imple- a laboratory, usually as tabletop experiments. Often large
mentations to ensure the constraints are not violated and Hilbert spaces can be encoded, e.g., the occupations of
the Fermi statistics remains intact. Symmetry-protection thousands of sites of an optical lattice by cold rubid-
strategies outlined in the previous section can, therefore, ium atoms, but there are only a few “knobs” that can
be of value in fermionic simulations with local mappings be turned, e.g., the optical lattice spacing, depth of the
to qubits. There has not been sufficient research on the per- potential, etc. Consequently, this approach could be suit-
formance of various bosonic and fermionic encodings in able for studies of universal properties or continuum limits
gauge-theory simulations but limited results have appeared of models in situations where the correlation lengths are
for select models and formulations [135,470]. The out- large. Research in analog quantum simulation of QFTs
come of such analyses will not only help with deciding requires understanding the underlying physics of the sim-
the optimal choice of formulation and encoding, but can ulator. For example, to determine if the capabilities of the
also lead to the development of new and better customized hardware can be matched with the features of the target
encodings for QFT applications. Hamiltonian, one needs to learn what the native or natu-
Finally, dedicated algorithms are needed for a range rally implementable interactions are in the simulator, and
of quantities of interest in high-energy physics, beyond what characteristics of the simulator can be tuned eas-
the scattering amplitudes, and such algorithms must spec- ily and what features are harder to modify. A dedicated
ify concretely the state-preparation and (partial) state- Appendix (Appendix H) reviews the status of the state-of-
tomography techniques that are efficient and tailored to the-art atomic, optical, molecular, and solid-state analog
the goal of the problem. For example, obtaining the platforms for quantum simulation, and their prospect for
hadron tensor for computing hadron’s structure functions simulating QFTs. Here, we focus on opportunities and
requires matrix elements of space-time separated quark- challenges of simulating quantum fields in an analog man-
level currents inside a proton and standard quantum algo- ner. Furthermore, the need for hybrid strategies, which
rithms for evaluating correlation functions exist, see, e.g., combine the benefits of the digital and analog schemes in
Ref. [471]. Nonetheless, questions regarding how to pre- one setting, is motivated, and is further elaborated in later
pare the hadron state to a given accuracy, and how to Appendices.
estimate the error associated with state preparation along Interacting scalar field theories, due to their equiva-
with that in correlator measurement, need to be fully lence to coupled many-body quantum harmonic oscillators
addressed. Besides the energy spectrum, particle-density appear more natural in systems that have an effective
distribution, and equal-time and out-of-time correlation description in terms of coupled harmonic oscillators, such
functions, entanglement measures such as entanglement as in superconducting circuits [472,473]. In fact, the first
entropy and entanglement spectrum will be of critical proposals for quantum simulating sine-Gordon models
value in our understanding the phases of matter in and have been put forward in recent years [474,475]. In con-
out of equilibrium, and how thermalization, fragmentation, trast, lattice gauge theories are coupled fermionic-bosonic
and hadronization occur in collider experiments. Research models that often do not have a simple harmonic-oscillator
is needed to assess the applicability and efficiency of representation for the gauge bosons. The Schwinger model
recently proposed methods in entanglement tomography, in the limit of large bosonic occupation can be approxi-
e.g., using random measurements and classical shadows mated by a spin-harmonic-oscillator system, for which an

027001-29
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

analog-simulation proposal has been developed in the con- error that can generally be bounded systematically. On the
text of trapped-ion systems with phonons as the bosonic other hand, analog quantum simulation can only simulate
degrees of freedom [256]. Similarly, the Abelian Higgs certain systems whose degrees of freedom and interactions
model is proposed to be studied using superconducting are similar to those of the simulator, but provides a more
microwave cavities that implement a boson-based varia- natural and resource-efficient approach to time evolution.
tional quantum algorithm [476]. Purely fermionic as well Combining the benefits of each mode of the simulator, one
as simple QLM formulations of the Schwinger model, can allow some degree of digitization in the simulation so
on the other hand, involve only interacting spins and that the engineering of the approximate dynamics is facil-
have received interest in the context of trapped ion [35, itated compared with the analog simulator. On the other
150,199], polar molecules [477], cold atoms [149,179– hand, it could be the case that the simulator offers access
181,203,478–481] including with Rydberg arrays [200], and control of certain degrees of freedom that can encode
and superconducting-circuit [152] simulators. In fact, the more naturally the degrees of freedom of the target the-
largest-scale analog simulations of the Schwinger model ory, or that there are intrinsic interactions among these
have been enabled in recent years within the QLM descrip- degrees of freedom that can allow a more extended set
tion [197]. The situation with non-Abelian LGTs is less of quantum gates to be devised. In this case, it is rea-
developed. The gauge degrees of freedom in Schwinger- sonable to devise digital simulations that are augmented
boson and loop-string-hadron formulations of SU(2) LGT by some analog building blocks. As an example, certain
admit a harmonic-oscillator description, but the number QFTs are shown to benefit from a hybrid approach in
of local oscillators and the complex nature of interactions trapped-ion platforms, where phonons can encode bosonic
among them do not have a natural analog in current simu- fields and participate in the dynamics actively [201,202].
lators. A few proposals exist for simulating the SU(2) LGT As another example, the quantum simulation of LGTs
in atomic simulators [151,203,204], including encoding can be digitized such that plaquette interactions involving
the non-Abelian Gauss laws as natural angular-momentum four- or higher-body terms arise as an effective inter-
conservation laws in atomic collisions [205], but these action when a series of two-body local interactions are
proposals have not yet been implemented in experiment. implemented with the use of mediating ancillary qubits
Going to dimensions higher than 1+1 D presents a [482,483]. Furthermore, higher-dimensional qudits might
bigger challenge, as both the electric and magnetic Hamil- prove useful [484] in encoding multicomponent fermions
tonian must be simultaneously implemented in the simu- or scalar fields [485,486]. A great deal of research is antic-
lator. When working in the irrep (electric-field) basis, the ipated to uncover the true power of quantum simulators
magnetic (or plaquette) Hamiltonian is complicated and for QFTs when such flexibility in the choice of simulation
requires interactions among at least 4 degrees of freedom, mode is allowed.
which are intrinsically harder to engineer in any of the cur- Another method to encode information in quantum com-
rent analog quantum simulators. This motivates the need puting is based on continuous variables (CVs), such as the
for dual-variable bases, in which the magnetic Hamilto- position and momentum of a particle or the quadratures
nian is more easily implemented, however, this leads to of an electromagnetic field. CV quantum computing was
more complex electric interactions that may not be natu- first proposed by Lloyd and Braunstein [487] who pre-
rally implementable in the simulator. Engineering a highly sented necessary and sufficient conditions for constructing
accurate minimal building block of an Abelian or non- a universal quantum computer over CVs. Unlike discrete-
Abelian LGT in 2+1 dimensions will be a milestone for variable quantum computing, the basic unit of information
the field in the coming years. Scaling the system while in CV quantum computing is a quantum system with
maintaining the fidelity will then be the next challenge an infinite-dimensional Hilbert space called qumode. CV
to overcome. To achieve this goal, extensive research is quantum computing is a natural platform to study contin-
needed in surveying and analyzing theoretically a range of uous quantum systems such as QFTs. One can also use
plausible analog-simulation platforms. This challenge also qumodes to encode qubits by using encoding schemes
motivates the need to search for the most optimal formula- such as the one proposed by Gottesman, Kitaev, and
tions of QFTs, noting the fact that depending on the digital Preskill [488] or using coherent states [489]. Addition-
and analog nature of the simulation, and the type of the ally, hybrid protocols that leverage the advantages of both
architecture used, different optimal formulations may be qubits and qumodes have been proposed [490]. Recent
found. experimental breakthroughs, such as the demonstration of
The engineering challenge associated with implement- quantum advantage using Gaussian boson sampling [491],
ing a continuous and fully analog evolution of gauge theo- and the introduction of a programmable photonic quan-
ries further motivates hybrid approaches to the simulation. tum computer [492], have sparked a growing interest in
One one hand, digital quantum computation allows simu- the development of CV quantum algorithms. The first CV
lating arbitrary local or semilocal Hamiltonians efficiently, quantum algorithm that studied the quantum simulation
providing an approximation to various unitaries with an of a QFT was proposed in Ref. [185] where scattering

027001-30
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

amplitudes in a scalar bosonic QFT were calculated, can be developed, tested, and optimized with existing
and later extended to scalar electrodynamics [186]. For NISQ devices. The general idea is to start with simple
the ground and excited states of a QFT, the quantum models in low dimensions (e.g., Z2 LGT, the Schwinger
imaginary-time evolution [493] algorithm adapted to CV model, etc.) and progressively increase the symmetries
substrates was recently proposed [494]. Energy levels in a and dimensions. An approximately similar sequence of
φ 4 QFT were calculated on a quantum simulator, and by models, sometimes called the “Kogut ladder” or “Kogut
developing similar algorithms, one can also study finite- sequence” [317,318] was proposed in the development
temperature systems [495]. More research is needed to of classical-computational methods, and it is clear that it
estimate the required resources for quantum simulations can also be of value in the quantum simulation era [132].
of QFTs using CV quantum hardware. Such investigations In fact, from the very first quantum simulation experi-
will need to include quantum error-correction schemes, a ment of a gauge theory in a small trapped-ion quantum
subject that is still in its infancy for CVs but early adap- computer in 2016 [187], the number of implementations
tations of known discrete-variable codes for CV quantum and benchmarks on quantum hardware continued to sub-
computing have emerged in recent years [496–498]. stantially grow. The simulations of the Schwinger model
More generally, in order for the analog and hybrid [40,115,162,188,192] have gradually improved compared
approaches to present a realistic path toward reliable simu- with the early demonstrations, and there exist now simula-
lations of the target theory, two complementary research tion results for small non-Abelian gauge theories as well,
areas must be developed. First, analytical or empirical including discrete DN group [189], and SU(2) [116,190]
analysis of the errors due to inaccurate engineered dynam- and SU(3) [118] LGT in 2+1 D. What these demonstra-
ics must be systematically performed but this generally is a tions show is that increasing the number of qubits in the
nontrivial task. It is also important to come up with error- hardware without comparable increase in gate fidelities is
correction and error-mitigation strategies that do not rely not of value as QFT circuits are not only qubit-resource
on the digitization of the evolution or qubit encodings. In intensive but also extremely deep.
the absence of such protocols, it is still conceivable that Analog quantum simulators have also been used to study
a set of criteria can be determined, such as the degree the dynamics of LGTs. These range from simulating small
of local and global symmetry violations or inconsisten- building blocks of a LGT, such as a link starting and ending
cies in different observables, to put into test the result at the fermion sites using two-component ultracold atoms
of the simulation. However, full confidence in the sim- in double-well potentials [194] and other variants [195]
ulation result may not be possible without independent for the Z2 LGT, and using interspecies spin-changing col-
implementations of the same problem on different plat- lisions in an atomic mixture in an optical lattice for the
forms, particularly as the simulation sizes grow beyond U(1) LGT [196]. Larger-scale demonstrations of gauge-
classical limits. Finally, an important task the HEP com- invariant dynamics in the spin-1/2 QLM of the U(1)
munity will take on in the coming years is to investi- LGT in 1+1 D have been made possible in defect-free
gate which problems are more suitable for implementa- arrays of bosonic atoms in an optical superlattice with
tion in analog simulators given potential inaccuracies and up to 71 sites [197], along with the first demonstration
errors. There may be features that require larger Hilbert of gauge-theory thermalization in this model [198]. Real-
spaces to be simulated than plausible in near-term digital ization of gauge theories in 2+1 D [179–184] and higher
devices, but are less prone to imperfections in the simu- dimensions and of the non-Abelian gauge-theory dynam-
lator. Those qualitative features are likely to be related to ics [151,184,203–205,499] in analog quantum simulators
studies of phases and phase transitions of matter. There- will mark an important next step but these yet need more
fore, there may be a unique opportunity for HEP physi- realistic proposals, as discussed before.
cists over the coming decade to identify the quantum The simulation experiments and implementations,
advantage case that may be enabled by analog or hybrid enabled either through access to cloud-based quantum pro-
simulators. cessors by companies or via university collaborations, are
a critical component of the quantum simulation program
for two reasons. First, they allow hands-on experience
5. Illuminating the path: implementation and with quantum hardware and fill the gap between theory
benchmark and algorithm and the simulation, and hence guide the
In the current NISQ era of computing, the number of developments toward more realistic and hardware-efficient
basic quantum computing units is limited and the gate proposals for experiment. Second, they allow the hardware
errors put constraints on the depth of the quantum cir- developers to become familiar with the unique problems
cuits. Similarly, analog simulators are limited in size and presented in QFTs and engage in a co-development pro-
quantum control, and face noise and coupling to the envi- cess where dedicated QFT simulators could perhaps be
ronment. Nevertheless, the building blocks of the uni- considered and designed. This co-design process is further
tary evolution necessary to study the simplest models discussed in a dedicated Appendix (see Appendix L).

027001-31
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

6. Combining classical computing with quantum initial state using conventional lattice-field-theory meth-
simulation ods, and encode this information in the quantum simulator
In the near or even far term, it may be beneficial to incor- to avoid the costly state-preparation step, as proposed and
porate classical-computing methods in quantum simula- tested in Refs. [207,208]. Unfortunately, if the initial state
tion so that quantum resources can be allocated more effi- is hard to compute classically, for example, in situations
ciently. This can enable simulations that otherwise would where a sign or signal-to-noise problem is present, the pro-
not be possible. For example, the spectrum of an interact- tocol is faced with the usual issues, but can otherwise be
ing quantum many-body system can be determined with valuable. Tensor-network-inspired state-preparation tech-
a quantum computer via known phase-estimation algo- niques can also be taken advantage of in this context.
rithms, which are nonetheless costly and not suited for the The final way in which classical resources will play an
present NISQ hardware. As an alternative, the variational important role is in refining the classical-quantum bound-
principle of quantum mechanics can be taken advantage of ary beyond which practical quantum advantage occurs in
to find the lowest-energy eigenvalues by inputting a para- HEP [504].
metric initial state and evolving it to nontrivial final states. More hybrid classical-quantum simulation protocols, in
Measuring the Hamiltonian matrix element and using clas- the spirit of those described in this section, can offer lower
sical optimizers can then allow putting rigorous and ideally quantum resource requirements compared with full quan-
tight upper bounds on energies. Variations of such vari- tum simulations, and can therefore speed up the progress
ational quantum algorithms (VQAs) [275,276], including in addressing the physics drives of this program.
quantum approximate optimization algorithms (QAOAs) In summary, a quantum simulation program in lattice-
[277], have been developed in recent years and have been field theory is starting to form. Its development over the
applied to a range of problems [278]. In the context of next decade will be guided by many insights from the
QFTs, the low-lying spectrum of lattice Schwinger model development and growth of the conventional lattice-field-
[40,115,188], and a first VQA applied to the SU(2) LGT theory program. It will rely on the existing and forthcom-
coupled to fermions in 1+1 D has appeared in recent years ing advancements in quantum algorithm and hardware,
[190]. Such a variational approach can also be adopted to but equally importantly on new proposals and algorithms
find good approximations to the eigenstates of the system, tailored to the unique features of QFTs, and particu-
which in turn can inform conventional lattice-field-theory larly gauge-field theories. QIS-literate quantum field the-
calculations that need good interpolating operators for the orists will be the key to these advances, but they can
states [210]. They have also been suggested to facilitate the accomplish much more by collaborations and exchanges
scattering problem on a quantum computer [206]. with the quantum information science and technology
Another area worth exploring is to determine if the con- community. As a result, quantum simulating QFTs will
ventional lattice-field-theory program can be accelerated be a highly interdisciplinary area of research in HEP
with quantum-computing routines. Here, questions that over the next decade, and is expected to combine the-
need to be answered include the following: Can quantum ory, algorithm, and hardware research in exciting new
processors be useful in enhancing importance sampling of ways.
quantum configurations, especially when sign or signal-
to-noise problems are encountered, or when a critical
slowing down halts the simulations toward the contin- APPENDIX G: UNDERLYING SIMULATIONS:
uum limit of lattice field theories? Can quantum platforms SIMULATING EFFECTIVE FIELD THEORIES
speed up inversion of poorly conditioned large matrices An effective field theory is a field theory that repro-
[500], enhance semidefinite programming for construc- duces a given underlying field theory in a particular
tion of optimal field interpolating operators or improving kinematic regime [505]. Examples are four-Fermi effec-
the nonperturbative bootstrap in conformal field theories tive theory [506], which describes weak interactions of
[501] (which can benefit from quantum advantages from elementary or composite particles with momentum trans-
Gibbs sampling [502,503]), or incorporate more economi- fers below the electroweak symmetry-breaking scales,
cally the factorial growth of the number of contributions in hadronic and nuclear effective theories [507–516], which
nuclear correlation functions? Initial investigations along describe interactions of hadrons and nucleons with low
these lines have started [209,210] and will likely continue momentum transfers, and soft-collinear effective the-
as the fault-tolerant quantum processors become closer to ory [15–18], which describes the interactions of partons
reality. collinear to each other, or the interactions of collinear
Another interesting path is to use the input from and soft partons. Since an EFT must reproduce the same
conventional lattice field theory to accelerate quan- physics as the underlying theory in the regime where it
tum simulation of QFTs. An example is to leave the applies, relevant information about short-distance physics
computationally demanding task of time evolution to the must still be contained in the EFT. The generation of EFTs
quantum simulator, but learn the density matrix of the follows a renormalization-group matching procedure with

027001-32
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

which irrelevant high-scale degrees of freedom are inte- the short-distance physics describing the initial production
grated out but their effect is encoded in the coefficients of the jets is contained in the Wilson coefficients. The full
of the generated operators and their scale dependence. dynamics of such cross sections are, therefore, described
These coefficients are typically called Wilson coefficients, by perturbatively calculable short-distance coefficients and
or in the context of hadronic and nuclear EFTs, com- nonperturbative long-distance matrix elements of SCET
monly named as low-energy constants (LECs). When the operators.
matching occurs at energy scales far above the hadronic The collinear dynamics within a given jet in SCET are
scale, these coefficients can be calculated perturbatively, the same as those of the full theory, albeit with a much
given that the strong interaction is asymptotically free. smaller dynamical range required. Thus, the same tech-
The LECs of the hadronic and nuclear EFTs can only be niques as those developed for the quantum simulation
constrained by matching to experiment or via a nonpertur- of full QCD using Hamiltonian lattice-gauge-theory tech-
bative evaluation within the underlying QCD theory using niques are applicable in this case, see Appendix F. The
the lattice-QCD method. important simplification, however, is that the dynamical
The separation of the overall dynamics into short- range has to be much smaller, therefore reducing signifi-
distance contributions and long-distance effects provides cantly the number of degrees of freedom required for a reli-
many advantages. First, short-distance contributions at able simulation. The first quantum simulation algorithms
high scale are often reliably calculable in perturbation the- for SCET dynamics has been developed in Ref. [222], and
ory. The remaining long-distance physics is then described will continue to be improved in the coming years.
by the dynamics of EFTs, which in many cases are easier to The soft dynamics of SCET, which is required to cal-
compute compared to the full theory. This is because when culate the nonperturbative soft matrix elements, is dra-
expressed in terms of the effective low-energy degrees of matically simpler compared with the dynamics of the full
freedom, the interactions are often simpler, and new sym- theory. This is because in the soft sector of SCET, collinear
metries may manifest themselves. Furthermore, the EFT degrees of freedom are integrated out, leaving only static
has to describe the dynamics only at long distances, and so color sources in the theory. Furthermore, interactions with
the EFT needs to be simulated over a much smaller energy soft fermions are power suppressed in SCET, such that
range than the full theory. the soft dynamics is described by those of a pure gauge
Since simulating the underlying QFTs at a wide range of theory in the presence of static Wilson lines. Since the
energies amounts to enormous computational cost, and that upper energy range of the soft theory is about 3 orders of
we do not know the valid QFTs of nature at a high scale magnitude below that required for a typical full theory sim-
(assuming these are QFTs), the renormalization-group ulation, a dynamical lattice-gauge-theory simulation of this
matching and the emergence of effective descriptions of soft theory requires a factor of 109 fewer lattice points than
interactions will continue to be a valuable approach in the corresponding simulation in the full theory. This leads
handling problems in HEP in the quantum simulation era. to a dramatically smaller resource requirement, such that
Quantum simulations have been studied for several dif- ab initio nonperturbative calculations of soft SCET matrix
ferent EFT applications: within SCET, for parton-shower elements seem feasible on quantum devices in the not-so-
physics, for parton distribution functions, and within distant future. A simplified version of this theory, namely a
hadronic and nuclear EFTs. These applications come with scalar field theory interacting with Wilson lines was stud-
distinct features, as described in this Appendix. ied in Ref. [222], along with necessary quantum circuits
and a small-scale simulation on an IBMQ quantum device.
The quantum simulation of a scalar field theory is by now
1. Simulations within soft-collinear effective theory very well studied [14,390,456–458], and the addition of
Most cross sections of interest at high-energy colliders Wilson lines is relatively straightforward. The progress in
such as the LHC are dominated by events that contain this problem is correlated with that in simulating lattice
only a relatively small number of hard jets. A jet of par- gauge theories in 3+1 D, which is discussed in Appendix F.
ticles is a collection of particles that have a small invariant
mass relative to one another. Jets, therefore, contain a col-
lection of energetic particles that are moving in the same 2. Simulating parton showers
direction (they are said to be collinear with respect to Collider events typically contain a very large number of
each other). Jets can interact with one another through the final-state particles. This can be traced back to the fact that
exchange of soft particles, which do not raise the invari- emissions can happen over a very large range in energies,
ant mass of any of the jets significantly. The long-distance giving rise to logarithmic enhancements of emission cross
dynamics that describes the interactions between collinear sections, and particle emission at low relative transverse
particles within a jet, and with soft particles that can medi- momenta does not involve a small coupling constant. For
ate long-range interactions between the jets, is described this reason, describing exclusive events with a large num-
by matrix elements in SCET [15–18]. On the other hand, ber of individual particles cannot be achieved using the full

027001-33
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

underlying theory. As already mentioned above, most of traditional lattice-QCD techniques. This is because lattice-
the particles in typical collider events are grouped into a QCD calculations are performed in Euclidean spacetime
relatively small number of jets, such that most emissions to avoid a sign problem. While several techniques have
are either collinear or soft relative to the particles produced been put forward to allow the calculation of PDFs in
in the underlying hard interaction. In the collinear limit, lattice QCD, with successful results in several cases [211–
one can show that the emission is dominated by single 215], a precise determination of PDFs, and particularly
amplitudes, such that quantum interference effects become the Bjorken-x-dependent distributions at small and large
subdominant and the emission can be described by a prob- values of x, remains challenging. Such determinations are
ability [7,10,12,223,224]. In the soft limit, emission can further complicated when PDFs are desired for atomic
still be described at leading order by an emission proba- nuclei used in high-energy collider experiments. Quan-
bility if one takes the limit of a large number of colors tum computers can, in principle, compute directly the
(Nc → ∞). The radiation of collinear and soft particles can forward light-cone matrix element relevant for PDFs using
therefore be described in a probabilistic manner, using a a first-principles lattice-QCD framework.
Markov-chain algorithm. Several proposals have been put forward in recent years
The probabilistic nature of a Markov-chain algorithm to demonstrate how PDFs can be accessed on a quantum
makes including quantum interference effects challenging. computer [216–220]. One important aspect of the matrix
Quantum interference effects that can be present are effects element relevant for PDFs is the gauge structure, in partic-
at subleading orders in 1/Nc , and interference between ular, the Wilson line required for gauge invariance. One
amplitudes with different intermediate particles and dif- may forego the complications in simulating the gauge
ferent internal kinematics. Initial studies have emerged structure and calculate the simpler hadronic tensor on a
in recent years to formulate the parton-shower descrip- quantum computer, then extract the PDF using perturba-
tion using quantum simulation methods. For example, a tive information for the partonic scattering [216]. This is
quantum algorithm has been developed in Ref. [225] to a similar strategy as to traditional extractions of parton
reproduce the regular parton shower while by comput- distributions from experimental measurements. However,
ing all possible amplitudes at the same time, it can be by using the results for a hadronic tensor obtained on a
constructed to include quantum interference effects. In par- quantum computer rather than using experimental mea-
ticular, it was shown in a toy model that this quantum surements, one has the ability to turn certain contributions
parton shower was able to include quantum interference off to make the extraction of particular PDFs easier. Alter-
effects arising from different intermediate particles with an natively, the Wilson line can be explicitly constructed
exponential improvement in efficiency compared to known using plaquette operators or fermion hopping terms, allow-
classical algorithms. ing for an estimate of the full quantum computation of the
More work is required to develop a full parton-shower PDF [217]. Finally, a PDF calculation in the Nambu–Jona-
algorithm for the SM, and to find ways to include the Lasinio (NJL) model was performed in Ref. [218] using
most relevant quantum interference effects, such as color a variational ansatz for the proton state, and following
interference. One can also look for other implementation Ref. [221] for the correlation function.
strategies of the quantum parton shower, such as those At this stage, it is not clear what the realistic
proposed in Refs. [517,518]. computational-resource requirements are for computing
PDFs and hadronic tensor to given accuracy, as the com-
plete algorithms, including that needed for preparation of
3. Simulating parton distribution functions hadronic states in QCD on a quantum computer, are either
Parton distribution functions are a crucial nonperturba- nonexisting or premature. Over the next decade, theoreti-
tive ingredient in any prediction for hadron colliders. In cal and algorithmic research will improve upon these initial
rough terms, they describe the probability to find a par- analyses.
ton with a given momentum fraction (or other properties
for generalized distributions) inside a hadron. PDFs can
be combined with partonic scattering processes to make 4. Simulations within hadronic and nuclear EFTs
predictions for hadronic scattering cross sections. Low-energy EFTs of nuclear interactions are important
PDFs are defined by the Fourier transform of a matrix for the HEP mission as they provide a consistent frame-
element of an operator containing two quark fields sep- work to describe the nuclear targets used in high-energy
arated in the lightlike direction, evaluated between a experiments. For example, accurate simulations of these
hadronic state. A Wilson line is required when forming the theories in systems of many nucleons are needed to com-
matrix elements to make the product of two fermion fields pute semiexclusive neutrino-nucleus cross sections that
at different locations gauge invariant. The fact that the two are needed for event analysis of long-baseline neutrino
quark fields are separated by a lightlike direction makes the experiments but are difficult to measure directly in the lab-
computations of this matrix element difficult when using oratory (see Appendix C for more details). They are also

027001-34
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

important to evaluate nuclear matrix elements required techniques, some of which can also be beneficial for
to analyze experiments looking for neutrinoless double- ground-state calculations. A crucial aspect to consider it
β decay [226,519], and for direct detection of potential the choice of encoding for the fermionic degrees of free-
dark-matter particles [227,228]. dom into qubits. Different mappings have, in fact, different
At low energies, these interactions contain both two- requirements in terms of the number of qubits for a given
and three-nucleon forces and are often nonlocal at higher number of nucleons A and of single-particle states N but
orders [520]. Simulation of these theories in a general also lead to spin representations of fermionic operators
single-particle basis, like the harmonic-oscillator basis, of different weight (the number of qubits they act non-
is structurally very similar to a quantum chemistry cal- trivially on). A conceptually simple and very common
culation, given the nonrelativistic nature of interactions. choice is the Jordan-Wigner mapping, which requires N
Therefore, quantum simulation techniques developed there qubits and O(N ) weight for fermionic operators [460]. The
can be ported, upon necessary modifications, to the nuclear large weight induced by this mapping is a major obsta-
case (see, e.g., Refs. [521–523] for recent reviews on the cle in the optimization of quantum simulation algorithms.
progress there). A few main differences are the presence of Alternative schemes have been proposed to overcome the
two additional fermionic species (to account for different difficulty, for example, the Bravyi-Kitaev mapping requir-
isospin components of the nucleon), the presence of three- ing N qubits but only O(log(N )) weight for fermionic
(and higher-) nucleon interactions, which in the most gen- operators [461], as well as recent semilocal mappings
eral case can lead to a Hamiltonian composed by O(N 6 ) with fixed weight [462–469]. Some work has already
distinct terms for a N -qubit system, as well as the presence started to explore the relative benefits of different fermionic
of pion-exchange interactions that in the static approxima- mappings in nuclear ground-state simulations [41] and
tion, lead to long-range two-nucleon potentials. This puts a an important direction in the future will be to extend
lower bound of the same order on the number of quantum these studies to dynamics as well as exploring alternative
gates needed to simulate real-time evolution, a requirement fermionic mappings that also feature error-correction prop-
to extract inelastic cross sections. Elastic cross sections erties at the expense of requiring a larger number of qubits
could be constrained by ground-state calculations alone [530] or auxiliary fermion methods [462,463,531].
and this could be performed with possibly much lower Another important research direction in the field is the
quantum resources using variational methods, at the price improvement of the approximation to the time-evolution
of requiring a number of measurements scaling as O(N 6 ) operator, which usually constitutes a dominant contribu-
(see, e.g., Ref. [524] for a review of quantum variational tion to the overall resource cost of algorithms that calculate
techniques). A substantial reduction in the number of mea- nuclear cross sections (see, e.g., Ref. [37]). A number
surements can be obtained in a number of ways: grouping of alternatives to the traditional Trotter-Suzuki decom-
operators into commuting families that can be measured at position have been proposed in the last few years, like
the same time (see, e.g., Ref. [525]), extending algorithms the Taylor-series method [443], quantum signal process-
that estimate the two-body fermionic reduced density ing [444], and quantum stochastic drift protocol (QDRIFT)
matrix (see, e.g., Ref. [526]) to the three-body reduced den- [532] among others. An initial comparison of the gate
sity matrix, exploring tensor factorization schemes (see, cost incurred by some of these in simulations of pionless
e.g., Ref. [527]) or adopting randomized-measurement EFT have been carried out in Ref. [38]. For more gen-
strategies [344,528]. To date, a number of variational cal- eral nuclear EFTs, which generate Hamiltonians with a
culations of small nuclei have been carried out on quantum large number of terms, randomized schemes like QDRIFT
hardware using simplified low-order interactions and small can provide important gains in gate requirements as their
model spaces with encouraging results [39–41]. Different cost does not depend directly on the number of terms in
methods for the direct extraction of elastic matrix elements the Hamiltonian but only on its norm. Finally, alterna-
have been also tested on the simple problem of deuteron tive methods for inelastic nuclear cross sections that do
photodisintegration [230]. not directly require the implementation of time evolution
Quantum simulations of ground-state properties of but instead use a quantum device to estimate appropri-
nuclear targets could become an important application for ate Chebyshev moments have been proposed [533]. As
future quantum technologies but the accuracy achieved some of the more advanced time-evolution schemes adopt
by classical methods for these problems sets a very high already a Chebyshev expansion [444], techniques of this
bar for quantum advantage [529]. Dynamical properties of kind could help in reducing the gate cost of simulations of
nuclei like inelastic cross sections are instead much more inclusive cross sections.
challenging to compute classically, especially for semiex- Finally, addressing questions regarding the composi-
clusive scattering in medium- and large-mass nuclei, and tion of the interior of neutron stars will not only shed
quantum simulations have the potential of being impact- light on our understanding of the phase diagram of strong
ful already on smaller-scale problems. Improving the interactions and the nature of the densest form of mat-
efficiency of real-time dynamics will require additional ter known in the cosmos, but also impacts the analysis

027001-35
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

of the gravitational-wave emission from merging neutron are reaching sufficient maturity to offer cloud-accessible
stars. Ab initio many-body calculations based in the under- hardware for analog or hybrid analog-digital simulations.
lying few-body hadronic interactions are computationally In the following, examples of analog-simulation platforms
challenging, but are crucial in discerning the role of non- with potential for simulating HEP models will be reviewed
nucleonic degrees of freedom in the equation of state of and the challenges ahead will be discussed.
neutron stars. Hadronic EFTs are generalized forms of
nuclear EFTs, and provide a systematic framework for
describing interactions among hadronic degrees of free- 1. Cold neutral atoms
dom, including baryons. These include chiral perturbation Systems of neutral atoms have been used as analog
theory [534–536] and baryon chiral effective field theory quantum simulators since the first production of quantum
[537–539]. Quantum computing the many-body problem degenerate Bose and Fermi gases [241–244], building on
with the hadronic EFT Hamiltonian can evade the intrin- fundamental advances in laser- and evaporative-cooling
sic sign problem in fermionic simulations, and is similar and leveraging the ability to trap and manipulate atoms
in nature to the computations described in this section for with the optical dipole force derived from intense laser
many-body nuclear systems. light. These devices, particularly utilizing purpose-built
Note that first-principles QCD-based simulation of optical lattices, have been used to probe quantum phases of
matter will still be necessary to complement this pro- many models in condensed-matter physics [541], explore
gram. First, they give direct access to the phase dia- fundamental phenomena in many-body dynamics [542],
gram of QCD without any assumption about the pres- and many other topics. A major advance in these sys-
ence of hadronic degrees of freedom. They can, there- tems came with the advent of the quantum gas microscope
fore, investigate the existence of the conjectured exotic (QGM) [245–247], which for the first time permitted very
quark phases [540]. Moreover, QCD-based determina- large systems of neutral atoms to be placed in a single
tion of the few-hadron interactions will be needed to many-body state through evaporative cooling, state prepa-
constrain the hadronic EFT at a range of energies and ration, and measurement. In many cases, QGMs can be
densities, which is experimentally challenging for short- used with atomic isotopes with both strong and weak col-
lived exotic hadronic states. A quantum-computing-based lisional effects as well as tunable by external fields [543].
lattice-QCD program in the few-hadron sector can, there- The single-particle dynamics is also highly configurable,
fore, be matched to a quantum-computing-based nuclear with a range of optical lattices with controllable geome-
and hypernuclear structure program via the EFTs (see, try, dynamical couplings, and Floquet engineering [248]
e.g., Ref. [40] for a first example), similar to the matching available. QGMs have since been constructed with both
program currently promoted using conventional methods bosonic and fermionic isotopes, and much of the benefit
[238–240]. The output of such efforts will subsequently of these systems has been devoted to quantum-degenerate-
impact research in HEP, including problems in the intensity gas phenomena where the indistinguishability of particles
and cosmic frontiers. plays a key role. Many of the systems studied in this
way exhibit complex quantum phenomena intractable with
classical numerical simulation, particularly those centered
APPENDIX H: SIMULATOR REQUIREMENTS:
on strongly correlated fermionic systems, and emerging
ANALOG SIMULATORS
studies on (non-Landau-Ginzburg) topological matter. To
Functional quantum simulators today can be found date, very few highly programmable platforms of this type
based on superconducting qubits, trapped ions, and neu- have emerged, and QGMs have been engineered physi-
tral atoms, but other platforms rooted in molecular, optical, cally from the ground up to support study of a particu-
and solid-state quantum systems are being developed and lar quantum simulation. New modalities [544], however,
explored for simulation application. Devices optimized for promise to accelerate the rate at which a given physical
programmable quantum simulation have been advanced, platform can be retasked onto new problems.
particularly those based on trapped-ion and neutral-atom More recently, cold atoms held in independently mov-
modalities. The requirements for HEP simulations place able optical tweezers and driven by laser light into Ryd-
a high priority on the development of simulators of large berg states have been developed as a platform for quan-
scale that exhibit good quantum coherence and high- tum information processing [545]. In these configurations,
quality readout, but perhaps more importantly a significant atoms are prepared singly near a motional ground state in
amount of control beyond what is customary in the past movable optical tweezers, and arranged into a defect-free
for quantum simulation of simpler spin systems. While initial register. Relying only on laser cooling, they achieve
access to superconducting qubit and trapped-ion systems an order-of-magnitude higher repetition rate than QGMs,
that operate in a digital mode have become available both and provide the run-time flexibility of nearly arbitrary
commercially and through DOE facilities and programs for geometry. Information is encoded in the internal state of
some time, it is only recently that several new modalities the atom, and strong interactions are introduced on demand

027001-36
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

using programmed optical coupling to an excited Ryd- degrees of freedom to the motional excitations of the ion
berg state. These flexible systems have advanced quickly crystals, a tunable long-range Ising interaction between
in the last 5 years, demonstrating long coherence times all pairsis achieved. This  Ising Hamiltonian engineered
and high state-resolved detection fidelities [546], elemen- is H = ij Jij σix σjx + i Bσiz , where the native spin-spin
tary gates [547,548], and recently the capability of moving interaction can be described by an approximate power law
atoms dynamically while preserving their coherence [549]. Jij ∼ J0 /|i − j |α , with 0.5  α  2, as well as an effective
In the simplest configurations, neutral-atom platforms of magnetic field that can be induced via applying Stark shifts
this type realize an effective
 spin interaction
 with a Hamil- on the ions. Trapped-ion systems have been used to sim-
tonian of the form H = ij Vij σiz σjz + i ( i σix + i σiz ), ulate spin models exhibiting such Ising Hamiltonians in
where i represents the strength of a local optical driv- up to a hundred spins [251,554,555]. These experiments
ing field coupling atoms into an excited Rydberg state, i take advantage of only a global pair of Raman laser beams
is its detuning from resonance, and Vij = C/|ri − rj |6 is to engineer the effective Hamiltonian, and because the
the strength of dipole-dipole interactions decaying quickly phonon is virtual in the process of coupling the spins, the
with distance for atoms positioned by tweezers at ri,j . scheme used is rather insensitive to the phonon occupation
The constant C is characteristic of the isotope and Ryd- [251]. On the other hand, being a second-order process in
berg state chosen in the implementation. In many cases, the strength of the native spin-phonon coupling, contribu-
systems of cold atoms can be understood approximately tions from the first-order processes present a source of error
from Rydberg-blockade physics, where it is assumed that in the Hamiltonian engineering. In practice, the coherence
no two atoms within a blockade radius (determined from times observed in the experiment, given this error source
the condition that Vij = i = j ) simultaneously transi- and the environmental and implementation noise, is of the
tion into the excited state. This is one simple basis, for order of a few inverse Ising coupling. Such simulations are,
example, for programming network-graph analysis [550] therefore, suitable for problems with fast time dynamics,
into an analog simulator using the arbitrarity of tweezer such as the evolution of quantum many-body systems after
positions ri —for “geometric” (unit-disk) graphs, nodes a quench [556–559].
correspond to atoms and edges represent closeness as Simulating QFTs requires more capabilities to be intro-
determined by the blockade distance. In the last year, duced to the scheme above. As a first requirement,
neutral atom devices utilizing Rydberg interactions have a similar level of control as in the digital trapped-
reached sufficient maturity to enable first demonstrations ion computers can be employed in the analog systems
of spin-liquid states [249], probed quench dynamics of sys- such that individual addressing of the ions with one
tems of unprecedented scale [250], extended modalities to or multiple pairs of Raman beams will be possible.
gates with shuttled qubits [549], and utilized long-ranged With the trapped-ion interaction graph being fully con-
and anisotropic interactions [551]. Neutral-atom Rydberg nected, a wide range of possible spin-spin Hamiltoni-
systems can represent information using multiple internal ans can be engineered in any dimension [560], includ-

states, and newly utilized isotopes of atoms have exhibited ing a Heisenberg Hamiltonian H = ij [Jij(xx) σix σjx +
extremely long coherence times due to their special level (yy) y y 
Jij σi σj + Jij(zz) σiz σjz ] + i Ji(1) σiz with independently
structure [546]. A number of commercial providers have
tunable coefficients, which is of relevance to simulating
already, or are near completion of, structured cloud access
the lattice Schwinger model [35]. Furthermore, the three-
to these platforms.  (3) + + +
Emerging platforms are likely to use modalities that spin Hamiltonians of the form H = ijk Jijk [σi σj σk +
− − −  (2) + −  (1) z
extend beyond fixed geometry, qubits, and also beyond σi σj σk ] + ij Jij σi σj + i Ji σi can, in principle,
single- and two-qubit gates [552], to dynamically recon- be generated with tunable coefficients, relevant for sim-
figurable connectivity [549], qudits and multilevel gate ulating simple QLMs among other models [199,252].
implementations [552], and efficient multiqubit entangling In fact, N -spin interactions have now been proposed in
gates [553]. trapped-ion systems with exciting possibilities for quan-
tum simulating various models of relevance to HEP, and
2. Trapped ions for simplifying digital circuits [253]. Moreover, spin-
In trapped-ion systems, atomic ions are confined phonon interactions have been taken advantage of in pro-
through electromagnetic fields, where the balance between posals for simulating the Schwinger model in an analog
their Coulomb repulsion and the external confinement fashion [256], and when combined with phonon-phonon
produces a well-defined crystal, typically in 1D or 2D gates can enable a hybrid analog-digital approach to sim-
in space. Each atomic ion stores an effective spin in ulating the same model and other coupled fermionic-
the same atomic states used as the most advanced fre- bosonic field theories [201,202], among other models [561,
quency standards, so their idle coherence is extremely long 562]. Further, the availability of multiple atomic states
[258]. The spins are manipulated through either optical or to encode higher-dimensional spin systems [254,255]
magnetic fields. By coupling off resonantly the spin adds more flexibility to the analog simulator trapped-ion

027001-37
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

toolbox. Many of these less conventional but more interactions to a target model requires a subtle co-
enabling schemes have yet to be tested in experiment, par- design process between hardware providers and simula-
ticularly when sizable simulations are desired. Later in this tion builders, see Appendix L. In analog simulators, this
section, the challenges and prospects for going beyond consists of matching both the target Hilbert space and
simple QFTs and toward gauge-theory simulations of the dynamical evolution by tailoring programming degrees-
Standard Model are discussed. of-freedom available in different devices. Since the native
architecture and the target system often do not share the
3. Other simulating platforms same fundamental symmetries or connectivity of interac-
tions, computing resources are inevitably lost to encoding
Other atomic, optical, molecular, and solid-state sim-
overheads. As newer modalities come online with effi-
ulating platforms are quickly advancing [257]. Among
cient multiqubit operations, new opportunities to mitigate
the notable platforms with potential prospects for simu-
these overheads will likely arise. Furthermore, in systems
lating models of relevance to HEP are (i) laser-cooled
that present fermionic and bosonic degrees of freedom,
polar molecules [563,564], which combine the strong elec-
as in certain cold-atom simulators, the issues with encod-
tric dipolar interaction with one to hundreds of internal
ing fermionic statistics and the need for low truncation of
states with transitions at convenient frequencies. Progress
bosonic modes can be circumvented.
is being made to address challenges such as cooling the
When it comes to simulating gauge-field theories of rel-
system to the many-body ground state and addressability
evance to the SM, the existing simulators are still far away
at the single-molecule level [565,566]. (ii) Cavity quantum
from presenting the essential capabilities and it is conceiv-
electrodynamics [567–570], which enables coupling of
able that the ultimate solution will be a hybrid approach:
distant atoms mediated by the exchange of the photon, or
a native and more versatile set of operations are used but
when viewed as many-body systems of cavity photons, the
the evolution is digitized to avoid the need for the chal-
strong photon-photon interactions mediated by atoms, with
lenging task of simultaneously applying an increasingly
the possibility of achieving programmable interactions
large number of terms in the local Hamiltonian of gauge-
and emergent geometries [177]. Future developments will
field theories toward higher dimensions and more complex
tackle the challenge of combining strong atom-light cou-
groups. While trapped-ion and cold-atom platforms have
pling with local tunability of the interaction of individual
been more extensively explored in the context of quantum
atoms with the cavity. (iii) Superconducting circuits [472,
simulation of gauge theories, it is yet to be known if the
473], which are the leading solid-state quantum simulation
other platforms briefly described in this section will pro-
platform, and are among the most geometrically flexi-
vide unique opportunities for this task. Intuitively, one may
ble simulators. They operate both in a digital mode, see
consider molecules as more natural candidates for encod-
Appendix I, and in an analog mode, by taking advantage
ing certain non-Abelian gauge theories, or take the super-
of linear and nonlinear couplings between their elements,
conducting cavities with hyperbolic interaction graphs as
i.e., cavity photons and artificial two- (multi)level atoms,
natural candidates for simulating physical models of rel-
to tailor interactions. Hyperbolic lattices using circuit QED
evance to HEP in curved spacetime. Nonetheless, such
[571,572] are being developed [176], which present a
connections must be carefully examined and developed.
potential platform for simulation of quantum effects in
While theoretical research in illuminating these questions
curved spaces [58]. (iv) Dopants in semiconductors such
must be pursued, these platforms will first need to con-
as in silicon, which provide fermionic degrees of free-
tinue to show their capability in simulating simpler and
dom encoded in conduction-band electrons that populate
more standard benchmark models before being considered
the array, and nuclear-spin degree of freedom of the dopant
seriously as candidates for HEP applications.
sites. Long-range Coulomb interactions makes possible the
quantum simulation of many-body problems modeled by
an extended Fermi-Hubbard Hamiltonian [573]. Effective APPENDIX I: SIMULATOR REQUIREMENTS:
control of tunable parameters in a donor-based simula- DIGITAL COMPUTERS
tors [574,575] is achieved through controlled placement of
Digital quantum computing implements algorithms as
donor atoms with atomic-scale precision [576,577], pre-
sequences of universal gate operations on the underly-
senting further opportunities for quantum simulation in
ing qubit architecture, based on, e.g., superconducting
these platforms in the upcoming years.
systems, cavity QED, neutral atoms, or trapped ions.
Once a native set of gates and connectivity constraints
4. Challenges and needs of HEP simulations are identified for different systems, high-level algorithms
While the systems described in this section are highly can be compiled and targeted for a variety of hardware.
versatile, and in some cases programmable, as exhib- Today, a number of platforms have reached commercial
ited by a large number of published proposals for novel maturity, and an ecosystem has developed around cloud
implementations of physics models, mapping the native access for these devices. Current technologies are limited

027001-38
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

today in scale below the 100-qubit level, and exhibit noise quantum processing units. Being semiconductor systems,
levels that prevent gate sequences in excess of typically SC QPUs can leverage extremely high-purity solid-state
tens of gates per qubit before the likelihood of error is of materials and sophisticated material-processing techniques
order unity. A number of metrics have been proposed for to produce QPUs with coherence times approximately
characterizing performance of digital quantum computing 100’s µs, high single-qubit, 99.8%–99.96%, and two-
devices, most prominently the quantum volume [578] and qubit, 98%–99.6%, gate fidelities [265], and to realize
the algorithmic qubit count [579]. Both are related to the chips of varied qubit counts from a few to close to a hun-
qubit count and the viable (single- and two-qubit-) gate dred. These QPUs offer the capability to execute proof-of-
depth, taking into account the resource overhead consumed concept quantum algorithms that move the boundaries of
by overcoming limitations of qubit connectivity. A wide producing and utilizing quantum entanglement on demand.
variety of mechanisms can be used to access commercial Such capabilities are available through commercial plat-
and noncommercial digital computing platforms, ranging forms and DOE testbeds and are increasingly being used
from cloud-service providers who work in concert with by scientists and engineers to explore novel quantum
commercial hardware vendors to government-sponsored simulations of physical systems, development of bench-
programs like DOE’s quantum testbeds, in which low- marking and error-mitigation protocols that feedback into
level hardware access to trapped-ion and superconducting improved SC QPU performance, and many other algorith-
qubit modalities can be made to introduce specialized con- mic implementations. Imperfections in control systems,
trol. Gate sequences can be specified in a growing number QPU design, electromagnetic environment, and materials
of high-level quantum computing languages, often in a are the culprits of suboptimal QPU operation performance.
hardware-agnostic way, and be provided to vendors for SC QPUs developers are constantly engaged in attempting
remote execution. While early software was developed both incremental progress and innovation to attack these
to run isolated single-instance algorithms, some support imperfections.
now exists for more sophisticated techniques, including Another closely related architecture, colloquially known
hybrid methods like VQE [275,276,278] and QAOA [277] as a 3D quantum processor, utilizes superconducting cav-
applied to gate-based algorithms. ities to store quantum information encoded in the vast
Trapped-ion quantum computers provide high num- number of levels (N ) available from each single-harmonic
bers of algorithmic qubits. The qubits exhibit substan- mode. Using a multitude of levels, as opposed to a qubit
tially long coherence times (minutes), and the single- that uses only two, effectively realizes 2 log(N ) all-to-
qubit and two-qubit gates can be operated with > 99.96% all connected qubits per harmonic mode. Moreover, these
and 98.5%–99.3% fidelities, respectively, and with state- 2 log(N ) “virtual” qubits are natively entangled, which
preparation and measurement error < 0.5%. Ion chains alleviates the necessity of many sequential conditional
with tens of functional qubits are achieved in current sys- gates to create a maximally entangled state as is required
tems, and importantly, are all mutually connected. This by the more traditional hardware. Beyond the direct encod-
feature reduces circuit depths as the qubits do not need to ing of quantum information, superconducting cavities are
be swapped and placed in proximity of each other to enable compatible with modern error-protection strategies and
a gate between them. The scaling of trapped-ion digi- highly efficient error-correction protocols, which makes
tal quantum computers will follow two predictable paths this platform a promising quantum computing platform.
[580]. First, the control of large ion crystals will be lim- Conveniently, the workhorse in this hardware implemen-
ited by the dense motional (phonon) modes that mediate tation, known as ancilla, is a conventional transmon qubit,
their interactions. This will demand that ion crystals be which allows one to control the quantum processor using
broken into spatially separated modules, with the quan- ordinary microwave pulses. Perhaps unintuitively, the
tum connections provided either by shuttling ions in space “limited” coherence time of the transmon does not factor
between the modules [260,581] or by photonic intercon- into the coherence time of the quantum hardware provided
nects between nodes [264]. Notably, this latter method the transmon coherence time exceeds the required gate
for modular expansion allows full connectivity, even to duration, which is easily achievable. The cavity coherence
scale. Second, as the number of qubits grows, the lim- time, however, does affect computational performance,
its of quantum gate fidelity will begin to determine the which is the motivation of the Superconducting Quantum
achievable circuit depths. At this point, it will become nec- Materials and Systems Research Center at Fermilab to use
essary to employ error-correction encoding to further the the typical superconducting radiofrequency cavities that
coherent quantum evolution. Error-correcting codes with are known for their world-record relaxation times. Ongo-
trapped ions are particularly efficient, owing to the full ing R&D is focusing on the demonstration of record-high
connectivity and natively low errors in trapped-ion systems coherence of cavities coupled to qubits. Multicell cavities
[268,582,583]. can implement powerful QPUs, with qudit-based and all-
Another digital quantum computing platform of par- to-all connection among multiple radiofrequency modes
ticular importance is Superconducting electronics-based with high-quality factors.

027001-39
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

While offering rich and novel opportunities, new pro- pushes the limits of classical simulation. It may be pos-
gramming paradigms, compilation, and transcription pro- sible to find algorithms that are more resilient to noise,
cesses will be necessary to understand how best to utilize along with features of the simulated theory that are more
these devices in HEP applications. For the purpose of sim- robust to imperfections in the implementation, so that use-
ulating gauge theories of the Standard Model, for example, ful predictions can be made with the NISQ device, espe-
it may be best to work with the formulations and encodings cially as targeted and effective noise-mitigation methods
that retain the locality of interactions, both among bosons are developed and applied. Furthermore, pure algorithmic
and among fermions and bosons, depending on the connec- developments can lead to theoretical scalings that some-
tivity pattern inherent to the hardware architecture used, in times prove pessimistic, while empirical scalings revealed
order to minimize costly swap operations. via NISQ-hardware implementation may turn out to be far
As the scale is increased in many platforms, consider- more promising. Since the quantum hardware design is at
able new opportunities will arise to implement encodings a stage where the winning candidates are still to be known,
of logical qubits that are resilient against gate, initializa- algorithm benchmarks given system architecture, connec-
tion, and readout errors. While proof-of-concept studies of tivity graph, degrees of freedom, and other parameters can
error correction with different encodings have been com- inform the next design choices.
pleted on various platforms, see, e.g., Refs. [267–272],
no system to date has sufficient resource to meaning-
fully utilize it in real-world algorithms. Through deep user 1. NISQ algorithms and strategies
interaction, co-design, algorithm innovation, and contin- There are a few promising approaches that can produce
ued improvement in digital-computing hardware, the field meaningful results even with shallow circuits and with-
is moving toward accelerating the timeline toward univer- out active error-correction sequences in the NISQ era. One
sal fault-tolerant quantum computation. Though it is likely well-formulated approach involves utilizing hybrid algo-
that analog quantum simulation and non-error-corrected rithms that divide classical and quantum resources such
algorithms represent the majority of near-term applica- that only steps that require probing a large combinato-
tions in HEP, an important thread of research will need rial space are executed on qubits. The variational quantum
to be devoted to the anticipation of larger-scale devices eigensolver [275,276] and quantum approximate optimiza-
with error-correction schemes. For the purpose of quantum tion algorithms [277] employ this hybrid approach. These
simulation, any model Hamiltonian with local and semilo- algorithms have been widely studied for their application
cal interactions can be decomposed into smaller units in combinatorial optimization, semidefinite programming,
of time evolution through the Trotter-Suzuki expansion and other contexts [278], including finding ground-state
[266] and efficiently decomposed into universal gate oper- energy of physical systems, see, e.g., Refs. [39,275,584].
ations. Other simulation algorithms have proven costly Similar techniques have gained attention for direct sim-
for present-day hardware but can be considered in the ulation of field theories, including finding the lowest-
fault-tolerant and large-scale era of quantum computing. lying energy spectra of low-dimensional Abelian and
non-Abelian lattice gauge theories [40,115,188,190] and
formulating costly scattering problems in terms of NISQ
APPENDIX J: SIMULATOR REQUIREMENTS:
variational algorithms [206].
SIMULATION IN THE NISQ ERA
Another approach to increasing the computational
Noisy intermediate-scale quantum era refers to an era power of shallow-circuit quantum algorithms for the NISQ
of quantum computing where noisy non-error-corrected hardware involves increasing the size of the operational
operations are performed on devices with approximately Hilbert space by using three-level systems, i.e., qutrits,
50–100 qubits [274]. Since the qubit and gate overhead or even higher-dimensional logic. This type of encoding
of error correction is substantial, and given that the gate increases the connectivity of the quantum network and
errors are not yet in the fault-tolerant regime, the circuit reduces the number of gates required for a given compu-
depths beyond a few hundreds to a few thousand gates tational task. As an example of the value of a four-level
will push the limits of sustained quantum coherence, and qudit, the four spin-isospin state of the nucleon (spin-
the noise affecting the computation must be mitigated in half neutron and proton) can be encoded in a qudit and
certain ways. While the full power of universal quantum the spin-dependent dynamics of nuclear systems can be
computing may not be reached for decades to come, it is more efficiently encoded and simulated, as demonstrated in
still imperative to not dismiss the possibilities provided Ref. [585]. This example also highlights another approach
by the NISQ hardware, and to take advantage of avail- to NISQ computing, that is to encode the portion of the
able devices with various capabilities and capacities. The system’s dynamics that is harder to track with classical
reason is that a quantum device capable of generating and methods on a quantum processor, and combine the results
manipulating an entangled state of a hundred qubits, even when possible with the components that can be evaluated
though for a short period of time and subject to noise, with more ease on a classical computer, in this case the

027001-40
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

evolution in position space of the nucleons. More examples number of gates in an algorithm, gate errors become
of such a hybrid classical-quantum simulation approach for dominant for longer circuits.
QFTs are provided in Appendix F 6, highlighting strategies There are several techniques that have been proposed
that are in play to maximally gain from limited quan- over the years to mitigate these errors. Readout errors can
tum power of NISQ hardware by porting more tractable be measured relatively easily by preparing a system in a
computations to classical hardware. given state and recording the measurement of this state.
The translation of a quantum algorithm into specific This results in a readout-error matrix. Note that given that
gates is not unique and various approaches can differ sig- the number of possible states is exponential in the number
nificantly in their actual quantum resource requirements. of qubits, the size of this matrix, and therefore the com-
For example, the original quantum algorithm for comput- puting resources to determine it, scale exponentially with
ing scattering amplitudes in a scalar QFT [13,14] uses the number of qubits. However, given this matrix, one can
an algorithm by Kitaev and Webb [586] for Gaussian- invert it using a variety of techniques [286], [286,584,590–
state preparation, which generates scattering wave packets. 596], where the last method is directly related to techniques
This algorithm requires polynomial qubits and gates, but developed for detector unfolding in HEP collider physics.
on NISQ devices, the required resources can be daunt- This corrects a given measurement on average. Recently a
ing, hence needing more optimal Gaussian-state genera- technique using active readout error correction, where each
tions [284]. Additionally, various techniques have been qubit to be measured is encoded into multiple qubits using
proposed to design minimal and maximally expressive repetition or hamming codes, have been developed [597,
quantum circuits for the NISQ era (e.g., Refs. [587,588]) 598]. This allows for readout-error correction for each
and to split large quantum circuits into small circuits exe- individual measurement. Examples of other approaches to
cutable on NISQ devices (e.g., Ref. [589]). Finally, gauge- readout errors are discussed in Refs. [599–601].
invariant simulation of gauge theories may save qubit Gate errors depend on the details of their implementa-
resources but locality of interactions may be lost. Many tion in the quantum simulator, and most of the commercial
digital simulations of lattice-gauge-theory simulations on systems available do not provide enough details to directly
the NISQ hardware to date [115,116,118,162,187,280] mitigate the errors introduced. One often distinguishes
have been enabled by fully (when possible) or partially between stochastic and coherent errors. Coherent errors
removing the redundant degrees of freedom (via appropri- preserve state purity. They are typically small miscalibra-
ate gauge transformation or by solving the Gauss law) or tions in control parameters. Coherent errors usually pro-
by imposing symmetries. This simplified simulation can duce similar errors in consecutive executions of a quantum
also reduce leakage to the unphysical sector of the gauge circuit and lead to a systematic bias in the output. Stochas-
theories due to algorithm and hardware errors. Taking full tic errors can be understood as either coherent errors with
advantage of such strategies for simulating the gauge the- randomly varying control parameters or as processes that
ories of the SM, while paying attention to the resulting entangle the system with its environment. Stochastic errors
time complexity of the simulation, will be essential in the can often be modeled as depolarizing noise. A method for
resource-limited era of quantum computing. converting coherent errors into incoherent errors is ran-
domized compiling [287–290]. The dominant gate error
is typically found in gates that introduce entanglement
between qubits, and a common universal gate introducing
2. Error mitigation for NISQ computing such an entanglement is the CNOT gate.
Given that in the NISQ era operations are noisy and the Among the common noise-mitigation techniques for
system is not error corrected, error mitigation is required CNOT errors (but which can be also used for other gate
to make sense of the measurements. There are two gen- errors) are zero-noise extrapolation (ZNE) and its vari-
eral types of errors that can affect a quantum simulation. ants [115,288,291–295]. ZNE magnifies the noise in the
The first is called readout error, and is responsible for the system either by changing the gate operation time or by
fact that the measurement of a qubit does not reproduce replacing each CNOT gate by a larger odd number of CNOT
the amplitude of the qubit before the measurement. The gates. Since applying two CNOT gates in direct succession
reason for such an error is often due to decoherence that amounts to an identity operation (CNOT2 = 1), this does
happens on the time scale required to measure a qubit. The not change the circuit in the noiseless limit, but it increases
second is called a gate error, and arises due to imperfect the amount of noise. This allows measurement of the result
implementation of quantum operators (gates). On current of the circuit at different noise levels, which can be used
publically accessible machines, readout errors typically to extrapolate the measurement to the noiseless limit. For
dominate, with error rates between 1 and 10%, and the the dominant depolarizing-error channel, one can prove
largest gate errors occur in entangling operators, such that this techniques reduces the noise level in a quantifi-
as the CNOT operation, with a typical error rate slightly able way, with the amount depending on the order of the
below 1%. However, since gate errors accumulate with the extrapolation used [293]. Other techniques have been

027001-41
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

proposed, which either require knowledge of the noise a low bar for entry to new users, with extensive docu-
model or use circuits to measure it [296–301]. Another mentation available online, including tutorials. Algorithms
interesting idea is to use the symmetries of the simulated that have been developed and integrated into HEP com-
system to control and mitigate errors [302,303]. An inter- putational efforts, such as lattice QCD, have been found
esting question that requires more study is whether gauge to be valuable to working with the annealing systems
invariance in gauge-field theories can be used to detect and [280,281]. Beyond optimizations for classical NP-hard
mitigate errors in the quantum simulators [147,148]. problems, a challenge currently faced in addressing classes
While noise mitigation is not a requirement that is of HEP problems is the dimensionality of their QUBO
specific to simulations for HEP applications, the HEP representations. The need for binary-string representations
community might be able to contribute to this impor- of complex numbers defining wave functions provides a
tant problem. For decades, high-energy physicists have challenge to the scale of problems that can be embed-
mastered the science of distinguishing tiny elusive sig- ded on the device. With next-generation annealers having
nals from dominant background noise, and of sheltering increased qubit connectivity and more qubits, larger prob-
sensitive particle detectors and readout instruments from lem instances can be addressed, but paths for tackling
environmental noise. It is conceivable that further noise- problems at scale (beyond preconditioning) remain to be
suppression techniques at the level of the hardware, and established.
noise-mitigation schemes at the level of algorithms, can
emerge from communications and collaborations among 4. Benchmarking for HEP
the two communities in the comping decade(s). In the context of quantum simulation of HEP, the prin-
Finally, while error mitigation will be important to lower ciples of locality and unitarity that played a crucial role in
the levels of noise of NISQ devices, and to obtain more the establishment of the SM also guarantees that dynam-
accurate results from them, error correction will likely be ics of QFTs of relevance to nature can be factorized into
needed to bring about the true power of quantum com- unitary building blocks that act on a finite Hilbert space
puters. This is the case especially for deep circuits, as independent of the size of the system. This is the central
the number of experimental measurements required for argument for quantum advantage in quantum simulation of
error mitigation grows exponentially with the circuit depth physical systems [266]. Developing, optimizing, and test-
[304–306]. ing these building blocks is a near-term task that can be
started with existing NISQ devices following the roadmap
3. Quantum annealers discussed in Sec. F. In this context, error mitigation, device
comparisons, and benchmarking are essential for making
Quantum annealers are nonuniversal quantum optimizer progress.
machines. Annealers [279] with more than 5000 qubits In particular, it is important to introduce metrics to
and associated simulators (with white noise) are accessi- assess the progress made and compare devices. This is
ble via cloud, and their performance is being benchmarked an essential aspect of the hardware and software develop-
across a wide range of problems, from commercial opti- ments that allows applications to be matched to platforms
mization problems, to finance, to applied mathematics, that best implement them. Indices have been proposed to
biology, materials, nuclear physics, and HEP. One niche compare mitigated real-time evolution with the exact Trot-
target for quantum annealers is to find optimal or near- ter evolution for different quantum devices [605]. Further-
optimal solutions to NP-hard problems, although their more, sophisticated methods such as cycle benchmarking
ultimate performance relative to classical competitors is [606] have been developed to characterize the systematic
unknown. In HEP research, the types of problems being errors of NISQ devices and applied to the quantum Ising
addressed include track reconstruction in the analysis of model [607] with a specific IBMQ machine.
collider data, e.g., Refs. [602–604] and real-time simula- The HEP community will continue to develop and apply
tions of elementary components required for QFTs [280, new NISQ-tailored simulation algorithms in the coming
281]. Annealers use time-dependent adiabatic switching to years. It would be interesting to see what the limit of NISQ
minimize the quadratic unconstrained binary optimization computing is for HEP applications and whether one can
(QUBO) cost function, or mappings to an Ising model, anticipate any quantum advantage in HEP in the NISQ era
constructed to address the problem of interest, includ- of quantum computing.
ing eigenstates and energies, real-time dynamics using the
Feynman clock algorithm, and problems that are NP-hard
APPENDIX K: SIMULATOR REQUIREMENTS:
(or NP-complete), such as max-cut problems.
SOFTWARE, AND COMPILERS
The quantum devices have extensive hardware controls
(e.g., controlling parameters of the annealing process) that Early investments by DOE HEP have enabled research
require higher-level expertise and allocations of device into algorithms and applications for the study of HEP
time to gain experience with. The D-Wave systems have problems on the NISQ devices. These early explorations

027001-42
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

have been critical to understanding how quantum comput- This will require the development of programming lan-
ers may be used to solve challenging physics problems. guages, potentially domain-specific languages, that can
Experiences with these early science explorations have readily express the discretization and complex interactions,
demonstrated the need to expand the programmability of such as the need to describe the coupling of fermions
quantum computing devices for testing new and creative and bosons and open quantum systems, provides key
quantum algorithms. building blocks, such as costly function evaluations using
arithmetic subroutines that are prevalent in quantum sim-
1. Current state of software and compilers ulating the dynamics of non-Abelian lattice gauge the-
ories, or large (sparse) matrix inversion for expediting
A variety of low-level quantum computing program-
Monte-Carlo-based routines present in the conventional
ming languages akin to assembler-level programming
lattice-gauge-theory program. Considering programmabil-
models in classical computing have been developed over
ity needs for digital simulations, it should be noted that
the past decade, and several of those have been used
there is a significant overlap with other science domains,
recently to program actual hardware devices. They are
including quantum chemistry and materials, that should be
low-level programming languages that work well for
exploited.
knowledgeable experts. Widely used software systems
In addition to programming languages and libraries,
include the IBM’s Qiskit [608], Xanadu’s Strawberry
the HEP research community would benefit from efficient
Fields [609], Google’s Cirq [610], and Microsoft’s Q#
compilers that can take the programs written in high-level
language [611].
programming languages to efficient low-level code that
Transpilers and early low-level compilers are being
can be run efficiently on quantum computing hardware.
developed to optimize the user-generated operations—
Interoperability of software is a key feature to ensure low
generally referred to as circuits—into a shorter and more
barriers for HEP users to access across different systems,
efficient set that can be translated to pulses and other
and avoid being restricted to a single, potentially propri-
fundamental operations needed to execute on quantum
etary software pathway. Open-source software and tools
hardware. Transpilation tools utilize multiple pathways to
will be essential for broad access, and tools that enable
optimize performance utilizing a wide variety of meth-
researchers to debug, verify, and validate quantum com-
ods, including graph optimization [612], unitary synthe-
puter programs and results will be needed. Last but not
sis [613], and ZX calculus [614] to name a few. Many
least, a hybrid computing model involving tightly cou-
of these tools also incorporate hardware-specific knowl-
pled classical and quantum hardware could be explored to
edge, such as topologies and error rates to deliver the
enable the usual offloading of special-purpose calculations
best performance possible. Full error correction is sim-
to a QPU [620,621].
ply not feasible for near-term quantum computers due to
Much of the discussion in this section has so far focused
limits in scale and the levels of noise. Nearly all sci-
on digital quantum computing, along the lines of the field’s
entific simulations require (mostly) nonautomated appli-
usage of classical computers. However, one could also
cation of error-mitigation approaches to achieve reliable
utilize quantum computers as well-controlled analog sys-
results [292,293,301,615,616].
tems, as discussed in Appendix H. For example, some
To ease the programmability for quantum computers,
recent work has utilized the cross-resonance effect of a
higher-level programming frameworks will be needed.
superconducting quantum computer to replace sequences
Some efforts have led to OpenFermion [617] for materials
of one- and two-qubit gate operations [622,623]. This
and chemical sciences and Pennylane [618] that is primar-
very different quantum computing paradigm will require
ily focused on machine learning. Within the DOE, efforts
that users to effectively engage across the quantum com-
are underway in the Office of Advanced Scientific Com-
puting stack, and to have access to software layers that
puting Research to develop a software toolkit that includes
generate pulses needed to drive the interactions of rele-
higher-level programming models [307]. No higher-level
vance to HEP applications, such as those of lattice gauge
programming models suitable for HEP are currently avail-
theories.
able, though homegrown tools are being developed at
Finally, one of the issues encountered during the adop-
various institutions.
tion of classical-computing resources for solving domain-
specific problems, such as lattice-gauge-theory simula-
2. HEP programming needs tions, was the diversity of paths and paradigms that
To advance HEP research on quantum computers, and required extensive effort to bring to a standard and uni-
enable the broader HEP community to readily partic- form state. Given the fast-evolving scene of quantum
ipate, models and approaches for programming quan- hardware technology, the computing paradigm might sub-
tum computers need to mature to a level in which stantially shift away from what may be considered standard
abstractions and library-based methods can be adopted today. Furthermore, the diversity of ideas and paths is a
to expedite programming and ensure portability [619]. necessity at such an early stage of developments.

027001-43
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

Nonetheless, the HEP community needs to arrive at a point In order for the co-design process to be effective, the
where consensus is reached on what the best simulation interaction between researchers within institutions that
strategies are to achieve the stated goals faster, and empha- deploy quantum hardware is essential. This allows syn-
size those paths and paradigms collectively so that the QIS ergies to be found that can lead to collaborations on the
community would recognize the concrete hardware and design of algorithms, protocols, and even devices, with
software needs of the HEP problems. immediate applications. Among successful examples in
the present day is the DOE-funded Advanced Quantum
Testbed [308] at the LBNL, where internal and exter-
APPENDIX L: QUANTUM ECOSYSTEM: nal collaborators and users are engaged in developing
HARDWARE CO-DESIGN AND ACCESSIBILITY benchmarking methods that offer new insights into noise
While it is, in principle, possible to construct gate-based processes and associated mitigation strategies in quantum
quantum computing algorithms without detailed knowl- hardware. This was done through an iterative process of
edge of the underlying hardware, the resource constraints protocol execution feeding back into the external theory
of today’s systems imply severe limitations on the scale of teams. To assist in this process, application-specific cir-
computations, and imply a need for hardware-aware effi- cuit synthesis, programming sequences, readout, feedback,
cient implementations and designs that take advantage of and noise mitigation were employed for maximizing cir-
a detailed understanding of system interactions and con- cuit depth and learning outcomes. Through these and other
nectivity, as well as error and decoherence mechanisms. interlocked collaborations and user interaction, the depth
This favors tightly coupled collaborative models for exe- and breadth of science that can be accomplished through
cuting science programs, in which hardware providers and novel technology can be advanced, enabling science that
domain experts in HEP applications co-develop scientific goes beyond cloud-based quantum computing resources.
agendas. This requirement also motivates the need for In addition to DOE programs at the national laborato-
accessible and adjustable devices such that HEP and other ries, several commercial hardware service providers now
domain scientists can use to benchmark algorithms and test offer in-house services to help design efficient hardware-
new ideas. cognizant implementations on demand [309–313]. This
will remove the wall between the users and cloud-based
services with preset features and will allow the needs of
1. Co-design models and special-purpose devices the domain scientists to be communicated directly to the
In the early days of the conventional lattice-field-theory hardware and software developers. At the university set-
program, research groups at a number of institutions ting, experimental efforts have long been in harmony with
around the world started to successfully design, build, theoretical proposals in quantum simulation, and there is
and run classical-computing hardware dedicated to QCD a direct relationship between strong theory groups and
simulations. One such development in the U.S. became strong experimental groups at the universities. Such a
the precursor to the so-called QCD-on-a-chip (QCDOC) model will potentially be the key to success in arrange-
machine, that via technology transfer to IBM became ments outside academia as well. For HEP applications,
the underlying design of successful Blue-Gene series of resources and opportunities in co-design efforts will need
high-performance supercomputers [624,625]. A similar to be directed toward systematic evaluation of the best
approach might prove valuable in the design and deploy- methods to simulate Hamiltonians of relevance to HEP
ment of quantum hardware for lattice-gauge-theory prob- efficiently and accurately, design and perform state prepa-
lems. In fact, such a co-design program is even more rele- ration and state tomography in strongly interacting QFTs,
vant in the quantum computing era as the simulating hard- understand the role of decoherence mechanisms, and how
ware is itself a physical system used for a physics applica- best to incorporate symmetries and gauge and scale invari-
tion. For example, as mentioned in Appendix H, Rydberg ance, or take advantage of them to mitigate or correct the
arrays offer the user a degree of flexibility in the design of errors.
the architecture of atom configurations hence model inter-
actions, trapped ions offer native bosonic degrees of free-
dom, i.e., phonon excitations that can be used to encode
more efficiently the bosonic fields and the interactions 2. Hardware-access model
beyond those possible with only qubits, and cavity QED In the rapidly evolving landscape at the interface
hyperbolic lattices can be engineered to simulate physics in between HEP and QIS, it is important for the HEP com-
curved geometries, along with many other examples. In the munity to identify optimal ways to have access to state-
next 5–10 years, scientific resources should be employed to of-the-art programmable quantum computers or quantum
explore a broad range of co-designed hardware schemes, simulation experiments. In the coming years, the ability
fostering a close collaboration between the HEP and the to access multiple platforms with a variety of architec-
QIS communities. tures would lead to rapid progress on target problems in

027001-44
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

HEP, as each platform would be most suitable for partic- process and potentially to special-purpose platforms. This
ular problems. As already mentioned, this access could should be accompanied by direct and broad access to
be through existing experimental setups, either located leading hardware and software technologies, and via a
in universities and national laboratories or by contract- community-approved mechanism that can unify, advo-
ing private companies. At the software and compiler front, cate for, and coordinate the accessibility needs of HEP
the most popular frameworks for programming univer- scientists.
sal quantum computers such as Qiskit, Cirq, and pyQuil
have so far been offered as free and open-source resources. APPENDIX M: QUANTUM ECOSYSTEM: SKILL
This has been critical for the HEP community, and will SETS AND WORKFORCE DEVELOPMENT
be important that this tradition continues as these pack-
ages become more advanced and gain commercial value. Quantum simulations of processes essential to HEP
By making quantum machines available to researchers, research objectives require a diverse and inclusive quan-
companies can ensure competitiveness in the future, which tum ready workforce with skills that extend significantly
is a practice now pursued by several countries, see, e.g., beyond those traditionally in HEP. This workforce will
Ref. [626–628]. Finally, it is also often necessary to run be distributed and collaboratively trained at universities,
relatively large simulations on classical computers, as national laboratories, and even some technology com-
numerical tests and benchmarks in the quantum simula- panies. The pipeline for recruitment should start ear-
tion problems is a necessity. It is conceivable that these lier, extending into high school. A complementary and
calculations start to require a non-negligible time at HPC robust portfolio of funding mechanisms, career develop-
facilities, and this requirement should be recognized ahead ment opportunities, career paths and mentoring will be
of time. required to create and sustain this workforce.
One successful model for how accessibility of resources
in HEP can be ensured is the USQCD-collaboration model. 1. Quantum workforce for HEP
USQCD [314] is a meta collaboration of the majority of Workforce development is a key component in advanc-
U.S. lattice gauge theorists. It has benefited from dedi- ing HEP’s quantum simulation objectives. Through the
cated hardware at Brookhaven National Laboratory, Fermi natural realignment of some of the existing workforce, and
National Accelerator Laboratory, and Thomas Jefferson the recruitment and training of junior scientists, the HEP
National Accelerator Facility, and provides access to these workforce is expected to evolve in a way to enable HEP
resources for the members via proposal solicitation and a objectives to be accomplished using quantum simulation.
rigorous review process. It also makes community-wide The intrinsically interdisciplinary nature of quantum sim-
bids for time on leadership-class computing facilities man- ulation for HEP requires growing or collaborating with
aged by the DOE, which has led to successful acquisition expertise in QIS, computer science, applied mathematics,
of computing time for large-scale lattice-QCD projects statistics, material science, nuclear physics, and more, with
over the years. One could envisage a meta collaboration scientists, engineers and developers residing in national
of U.S. HEP QIS theorists whose main goal would be to laboratories, universities, and technology companies. At
strengthen individual efforts and coordinate their access this time, this growth and engagement has been under-
to different hardware platforms available at national labs way in HEP for approximately 5 years, and it will need
or commercial sites. It would also facilitate communica- to continue and be further enhanced to meet HEP scientific
tions between people developing QC hardware and the objectives during the next decade and beyond.
needs of the theorists developing algorithms and software The growth of the quantum-ready HEP workforce is
for HEP applications. USQCD has also been successful in occurring in the context of significantly increased national
bringing up a collaborative algorithm-development team quantum information science and technology research and
feeding into independent and basically competing individ- development to meet broad-based present and future needs
ual groups. It can be thought of as a “virtual U.S. lab” of the nation [629]. When mature, the associated quan-
that could be emulated for the QIS community. It is also tum economy is expected to be comparable to the silicon
important to identify the most pressing needs of the poten- economy, and with a workforce of scale to enable a robust
tial users. For these reasons, the communication between and efficient pipeline from basic research to commodity
the HEP researchers and experimentalists developing new devices in quantum computing, communication, sensing,
computational platforms need to be enhanced, and these and more. It is important to recognize that the growth of the
communications can be strengthened by a unifying voice quantum ready HEP workforce, with skill sets necessary
within the HEP community. to meet HEP objectives, will be of significant benefit to
In summary, in the quantum era of computing, progress the national quantum activity. Historically, HEP scientists
in quantum simulation of HEP problems will be fostered have contributed to, or moved into, other areas of research,
by close collaborations among HEP and QIS scientists such as computing, “big data,” and device fabrication, and
with a range of expertise, leading to a valuable co-design have shaped the development of those areas in substantial

027001-45
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

ways. The same is anticipated in the emerging quantum instance, through courses in HEP, nuclear physics, and
era. condensed-matter physics, would directly benefit efforts in
The present “Snowmass” process is the first in which HEP quantum simulation. It would also be beneficial for
QIS is being considered an integral part of HEP research universities to integrate quantum circuit design between
and development, and consequently, estimates of the physics, engineering, and computer science. Such moves
resources and training required to meet the HEP QIS needs are already underway, but the inclusion of HEP relevant
are less sophisticated than in the more mature areas. This offerings would be beneficial.
next period is one of research and development in basic sci- Given the scope of the anticipated quantum ecosystem,
ence through to advanced technologies. This includes con- the pipeline for quantum education and training should
tinued calibration, development, resource estimation, and begin before students reach university. It would be bene-
engagement, including in the development of a quantum ficial for introductory QIS to become part of high-school-
ready HEP workforce with successful career opportunities. level science requirements. The QuarkNet program [630]
for developing the future technical workforce stands as
an existing example of how the community could inte-
2. Required skill sets and the role of universities, grate HEP QIS at this level. Increased opportunities for
national labs, and private sector high-school students to work with scientists at universities
The skills required for quantum simulations of processes and national laboratories are expected to be beneficial to
essential to the HEP mission are diverse [316]. Experi- quantum simulation efforts, including in HEP.
ences from the last decades of workforce development in The National Quantum Initiative (NQI) centers [315]
lattice QCD, event generators, and other HPC-intensive and QuantiSED programs effectively support the devel-
research areas, have provided models of engagement that opment of junior scientists, including postdoctoral fellows
are expected to have applicability. Collaborations of scien- and graduate students. While Covid-19 has provided a sub-
tists, engineers, and developers will be required to coalesce stantial limitation to in-person meetings, workshops, and
the skill sets required to accomplish quantum simula- collaborations, the regular meetings and activities have
tion objectives. These skills include HEP phenomenol- proven valuable in education and training. More QIS train-
ogy, quantum field theory, and quantum mechanics, lattice ing and educational opportunities at all career levels are
field theory, HPC, statistical analysis, experimental design required. For example, there will be utility in support-
and optimization, machine learning and artificial intelli- ing HEP-specific summer schools in quantum simulation,
gence, software stack development, quantum and classical- like the recently started Quantum Computing Internship
computing algorithms, quantum circuit design, implemen- for Physics Undergraduates Program at FNAL, alongside
tation and optimization, and more. Integration with quan- the more general schools in QIS that are organized by Los
tum hardware development through HEP co-design efforts Alamos National Laboratory. Programs analogous to the
will further broaden this skill set. The education and “lattice-QCD Hackathons” are anticipated to have a future
training programs for these skills are distributed within role in training for quantum simulation research. Named
universities, national laboratories, and the private sector. fellowships are used successfully at labs and universities
Therefore, coordination between these sectors is required, to attract talent at the postdoctoral and Ph.D. levels for
and all have an important role to play. For instance, one strategic purposes, and such fellowships for HEP quantum
expects universities to provide basic to advanced education simulation would be of value.
in physics, including quantum mechanics, quantum field
theory, and particle theory, along with basic to advanced
education in statistics and applied mathematics. National
Laboratories continue to develop the talent produced by 3. Career development and mentoring
universities and technology companies, and can assem- In order to meet the needs of the nation at all scales,
ble significant teams to address mission-specific objec- including those for HEP quantum simulation, career devel-
tives. Technology companies, in part, perform focused opment and mentoring has to be available at all stages and
research and development to produce commodity products. levels, for those so wishing. It is important to emphasize
For quantum simulation, this includes developing quan- that this area of research did not exist until recently, and
tum computers and quantum devices, developing robust further it has a different required skill set for the participat-
software stacks, and application programming interfaces. ing workforce. Until recently, the workforce in this area
These educational and training pipelines need to be has come exclusively from transitions from other areas,
both strengthened and expanded in the area of QIS. While and the first generation of QIS-HEP scientists is currently
physics departments at universities specialize in educa- being trained. Therefore, the landscape of research in HEP
tion in physics, both experimental and theoretical, an QIS has been changing quite rapidly, but is anticipated
increase in the advanced offerings in broadly defined quan- to stabilize during the next decade. There is only anecdo-
tum field theory and quantum many-body systems, for tal information available regarding career paths, how one

027001-46
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

is best positioned for a career in QIS, what courses stu- graduates and postdocs leaving HEP research for a career
dents should take, and so forth. As universities introduce in the private sector. The HEP experts in universities are
new courses, there will be differences between offerings at sure to be attracted to technology companies, and efforts
different institutions. This makes mentoring by members should be undertaken to future proof the HEP program for
of our community, at all levels, crucial, as is engagement the long-term health of the community.
with those in the national laboratories, the business sec- Finally, it is important to note that quantum comput-
tor, the legal sector, etc. This needs to be addressed during ing and quantum simulation will always involve a physical
the upcoming period. Programs such as the National Sci- system and it is unlikely that physicists, for the foresee-
ence Foundation’s ExpandQISE [631] will be essential in able future, treat such systems as a black box to solve
empowering smaller and less-involved institutions with the problems. Furthermore, even in the era of universal fault-
QIS research. tolerant quantum computing, physics insight and guidance
The next period will be one of experimentation to iden- are going to be essential in taking advantage of the com-
tify the types of mentoring and career guidance that are puting resources in the most optimal way, as has been the
effective. Importantly, one type of institution will no longer case with classical algorithms and computational meth-
be sufficient to impart all of the skills that are required, and ods. As a result, quantum computational physicists will be
collaborative mentoring should be explored and assessed. essential in developing application-specific algorithms and
Therefore, new and creative ways of educating and train- software. Academic institutions would need to recognize
ing junior, midcareer, and senior scientists, engineers, and the value they bring to the physics community in general,
developers are to be encouraged and explored. Impor- and the HEP field in particular, and not treat them as sole
tantly, this exploration should include drawing talent from computing practitioners, rather as physicists who deserve
previously underconsidered groups. (full or joint) permanent positions in physics. Along the
In designing and exploring potentially new avenues of same lines, R&D will continue to be a major part of quan-
developing this needed HEP QIS workforce, flexibility tum computing research in HEP, and physics results with
should be considered so that underrepresented sectors of impact on the HEP theoretical and experimental programs
the population have unimpeded and equal access to QIS may not materialize within the career span of a graduate
education and training, unlike previous generations, and student, a post-doc, or even a junior faculty. The value of
are welcomed into the HEP community. For this inclu- this endeavor may be established in the long run, hence it is
sive and diverse community to thrive, its members must important to recognize the nature of this research, and not
be supportive of each other and of the broader scien- penalize (junior) scientists that invest in developing novel
tific goals, and behave respectfully, ethically, and honestly. frameworks that can enable future physics achievements.
Codes of Conduct are now an integral part of formal In summary, the HEP community is starting to develop
collaborations in the community, and will continue to a quantum simulation workforce in order to address impor-
be important in future efforts. HEP should participate in tant objectives of the HEP scientific mission. For a diverse
the on-going national-level discussions on this crucially and inclusive quantum ready HEP workforce of scientists,
important subject. engineers, and developers to be at the forefront of quan-
One of the lessons learned from the recent decades of tum simulation, close collaborations between universities,
SciDAC [632] and HPC-oriented projects in HEP is in national laboratories, and technology companies are criti-
regard to the career path of senior researchers who start cal, see also Appendix N. The types of resources and the
as domain scientists and evolve into specialist algorithm nature of engagements that are anticipated to be required
designers and coders for HEP. This set of talented and for this to be successfully accomplished were outlined in
essential scientists and developers will typically leave this Appendix.
the community to assume positions in the private sector.
This impacts the workforce, while building connections
APPENDIX N: QUANTUM ECOSYSTEM:
with the private sector. Without identifying more desirable
ACADEMIA, GOVERNMENT, INDUSTRY
career paths within HEP, this situation may repeat itself in
PARTNERSHIP
the HEP quantum simulation era.
Joint positions between laboratories and universities Quantum simulation in HEP is a unique research disci-
have been successful in sustaining the HEP workforce. In pline that cannot grow in isolation. When algorithm, soft-
contrast, the use of strategic bridge positions to strengthen ware, and hardware developments are concerned, there are
the HEP workforce has been underutilized. As there is a multiple players in the field, from universities to national
strategic need to increase the workforce in quantum simu- laboratories to private companies, each contributing to
lation for HEP, consideration should be given to using joint the advancements according to the organizations’ strategic
and bridge positions at universities at the Assistant Pro- agendas. It is important for HEP scientists to acknowl-
fessor level (or higher) to attract talent into the program, edge this entangled web of contributors, and to form
and also to address the expected in-balance between Ph.D. partnerships that are long lasting and mutually beneficial.

027001-47
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

1. The role of national laboratories in advancing to be at the forefront of quantum simulation, engage-
quantum simulation ment with technology companies is critical, and will be
HEP has a long-standing history of driving develop- mutually beneficial. This engagement spans areas such
ments in classical HPC, an example of which is the as quantum hardware, operating systems, and application
emergence of IBM’s BG/L,P,Q supercomputers start- programming interfaces.
ing from the collaborative efforts of Columbia Univer- A somewhat novel aspect of HEP QIS is the need to
sity, RIKEN, and and Brookhaven National Laboratory address IP rights within collaborations, which will involve
(BNL). An important outcome of this co-design effort and legally binding agreements. This represents somewhat of a
HPC testbed(s) were the technical skills acquired through “different culture” for many in the HEP community, and
“hands-on” experience(s). Similar collaborative efforts are perhaps this may be a conceptual hurdle that has to be
emerging in HEP in the areas of machine learning and overcome as it does collide with the basic concepts of
artificial intelligence, as well as quantum computing, the open science. It has to be the case that all developments
results of which will benefit many scientific and technol- of importance to HEP research from engagements with
ogy communities. technology companies are future proofed. That is to say
The QuantiSED program(s) through DOE HEP that that any advance that enables accomplishing one or more
successfully supports QIS in HEP includes integrated col- objectives of the HEP mission must be able to reside in
laborative efforts including FNAL, LANL, BNL, LBNL the community and not be lost behind an IP barrier if a
with universities and technology companies, and brings company decides that this line of research is no longer a
together experts from HEP and other areas, such as com- priority.
puter science, atomic-molecular-optical physics, machine
learning, nuclear physics, condensed matter, and QIS, with
a significant effort related to quantum simulation. Research 3. Engagement with other domain sciences,
groups at the laboratories have skills related to, and are engineering, and computing
focused on, accomplishing the objectives of the laboratory. The types of problems that require quantum simulation
The QuantiSED programs provide a strategic supplement in HEP are similar in form to those in the other domain
to these skills through a closer coupling to outside experts. sciences. For example, nonequilibrium dynamics in quan-
It is expected that such activities will be essential for fur- tum many-body systems is of interest in QIS, condensed
ther developing the HEP QIS workforce going forward. matter, nuclear physics, fluid dynamics, fusion, biology,
It is also expected that the QIS expertise at the national as well as HEP. It is also conceivable that the techniques
laboratories with a HEP footprint to continue to increase that are being developed for the simulation of QFTs may
significantly during the next period to meet HEP objec- be of relevance for quantum sensing and quantum com-
tives. A similar evolution is anticipated in university-based munications. For example, the entanglement structure of
HEP groups, but in a way that complements the growth at quantum fields and their response to external probes may
the laboratories. This is, in part, to provide an education be harnessed in designing more efficient quantum sensors.
beneficial to the laboratories. Nonetheless, more research is needed to establish such
For HPC-centric HEP objectives, such as those advanc- connections in the coming years [633]. Workshops, meet-
ing lattice QCD and event generators, the SciDAC pro- ings, and collaborations, as are being enabled by the NQI
gram has played a critical role, bringing together domain centers, are crucial to optimal exchange of ideas and trans-
scientists and experts in applied mathematics, computer lations between languages and notations of these different
science, statistics and more through the SciDAC Insti- communities. In order for the communities to not have to
tutes. These valuable projects are typically jointly sup- reinvent the wheel, it is important that areas of overlap are
ported by DOE HEP and ASCR. Analogous opportunities identified and developments are being transported across
for advancing quantum simulation to meet HEP scientific seemingly disjoint disciplines. For example, lessons can
objectives would provide a valuable resource. be learned from how quantum information impacted and
advanced studies of condensed-matter systems over the
past decade, since it is likely that the HEP research can
2. Engagement with technology companies be similarly invigorated by the QIS concepts and tools.
Finally, HEP will likely be instrumental in motivating and
A robust engagement between universities, national lab- advancing QIS research, thus an integrated HEP QIS effort
oratories, and technology companies is essential. There will be mutually beneficial in quantum simulation.
is significant and growing expertise already in the pri-
vate sector in all areas of quantum. To minimize the
time from development of an idea, prototyping and sub-
APPENDIX O: COMMUNITY ENDORSEMENT
sequent integration into a commodity device, close ties
with technology companies are vital. In order for HEP 1. M. Sohaib Alam

027001-48
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

Quantum Artificial Intelligence Laboratory for Artificial Intelligence and Fundamental Inter-
(QuAIL), NASA Ames Research Center, Moffett actions, Massachusetts Institute of Technology, 77
Field, California, 94035, USA Massachusetts Avenue, Cambridge, Massachusetts
USRA Research Institute for Advanced Com- 02139, USA
puter Science (RIACS), Mountain View, California, 16. Bryce Gadway
94043, USA Department of Physics, University of Illinois,
2. Mohsen Bagherimehra 1110 W. Green Street, Urbana, Illinois 61801, USA
Chemical Physics Theory Group, Department of 17. Daniel González-Cuadra
Chemistry, University of Toronto, Ontario, Canada Institute for Theoretical Physics, University of
Institute for Quantum Science and Technology, Innsbruck, 6020 Innsbruck, Austria
University of Calgary, Alberta, Canada Institute for Quantum Optics and Quantum Infor-
3. Aniruddha Bapat mation of the Austrian Academy of Sciences, 6020
Physics Division, Lawrence Berkeley National Innsbruck, Austria
Laboratory, Berkeley, California 94720, USA 18. Alexey V. Gorshkov
4. Douglas H. Beck Joint Quantum Institute and Joint Center for
Department of Physics and Illinois Quantum Quantum, Information and Computer Science,
Information Science and Technology (IQUIST) Cen- NIST/University of Maryland, College Park, Mary-
ter, University of Illinois, 1110 W. Green Street, land 20742, USA
Urbana, Illinois 61801, USA 19. Erik Gustafson
5. Paulo F. Bedaque Fermi National Accelerator Laboratory, Batavia,
Department of Physics and Maryland Center for Illinois 60510, USA
Fundamental Physics, University of Maryland, Col- 20. Mohmmad Hafezi
lege Park, Maryland 20742, USA Joint Quantum Institute, University of Maryland,
6. David Berenstein College Park, Maryland 20742, USA
Department of Physics, University of California, 21. Jad C. Halimeh
Santa Barbara, California 93106, USA Department of Physics and Arnold Sommerfeld
7. Joseph Carlson Center for Theoretical Physics (ASC), Ludwig-
Theoretical Division, Los Alamos National Lab, Maximilians-Universität München, Theresienstraße
New Mexico 87545, USA 37, D-80333 München, Germany
8. Alessio Celi Munich Center for Quantum Science and Tech-
Departament de Física, Universitat Autònoma de nology (MCQST), Schellingstraße 4, D-80799
Barcelona, E-08193 Bellaterra, Spain München, Germany
9. Shailesh Chandrasekharan 22. Aram Harrow
Department of Physics, Duke University, Durham Center for Theoretical Physics, Massachusetts
North Carolina 27708-0305, USA Institute of Technology, Cambridge, Massachusetts
10. Anthony N. Ciavarella 02139, USA
InQubator for Quantum Simulation (IQuS), 23. Philipp Hauke
Department of Physics, University of Washington, INO-CNR BEC Center and Department of
Seattle, Washington 98195, USA Physics, University of Trento, Via Sommarive 14,
11. Eleanor Crane I-38123 Trento, Italy
Cambridge Quantum Deutschland GmbH, 24. Florian Herren
Leopoldstrasse 180, 80804 Munich, Germany Fermi National Accelerator Laboratory, Batavia,
12. Marcello Dalmonte Illinois 60510, USA
The Abdus Salam International Centre for The- 25. Ciaran Hughes
oretical Physics (ICTP), strada Costiera 11, 34151 Theoretical Physics Department, CERN, 1211
Trieste, Italy Geneva 23, Switzerland
13. Andrea Delgado 26. Travis S. Humble
Oak Ridge National Laboratory, Oak Ridge, Ten- Oak Ridge National Laboratory, Oak Ridge, Ten-
nessee, USA nessee, USA
14. Xi Dong 27. Joshua Isaacson
Department of Physics, University of California, Fermi National Accelerator Laboratory, Batavia,
Santa Barbara, California 93106, USA Illinois 60510, USA
15. Lena Funcke 28. Karl Jansen
Center for Theoretical Physics, Co-Design Cen- Deutsches Elektronen-Synchrotron DESY, Plata-
ter for Quantum Advantage, and NSF AI Institute nenallee 6, 15738 Zeuthen, Germany

027001-49
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

29. Natalie Klco Department of Physics and Astronomy, Rutgers

Institute for Quantum Information and Matter University, Piscataway, New Jersey 08854-8019,
and Walter Burke Institute for Theoretical Physics, USA
California Institute of Technology, 43. Alexander Schuckert
Pasadena California 91125, USA Cambridge Quantum Deutschland GmbH, Leop-
30. Michael Kreshchuk oldstrasse 180, 80804 Munich, Germany
Physics Division, Lawrence Berkeley National 44. Hersh Singh
Laboratory, Berkeley, California 94720, USA InQubator for Quantum Simulation (IQuS),
31. Andreas Kronfeld Department of Physics, University of Washington,
Fermi National Accelerator Laboratory, Batavia, Seattle, Washington 98195-1550, USA
Illinois 60510, USA 45. Jesse Stryker
32. Doga Murat Kurkcuoglu Department of Physics, and Maryland Center
Superconducting Quantum Materials and Sys- for Fundamental Physics, University of Maryland,
tems Center (SQMS), Fermi National Accelerator College Park, Maryland 20742, USA
Laboratory, Batavia, Illinois 60510, USA 46. Federica Maria Surace
33. Nima Lashkari Department of Physics and Institute for Quan-
Department of Physics and Astronomy, Purdue tum Information and Matter, California Institute of
University, West Lafayette, Indiana 47907, USA Technology, Pasadena, California 91125, USA
34. Randy Lewis 47. Shan-Wen Tsai
Department of Physics and Astronomy, York Uni- Department of Physics and Astronomy, Univer-
versity, Toronto, Ontario, M3J 1P3, Canada sity of California, Riverside, California 92521, USA
35. Andy C. Y. Li 48. James P. Vary
Fermi National Accelerator Laboratory, Batavia, Department of Physics and Astronomy, Iowa
Illinois 60510, USA State University, Ames, Iowa USA
36. Norbert M. Linke 49. Wei Xue
Joint Quantum Institute and Department of Department of Physics, University of Florida,
Physics, University of Maryland, College Park, Gainesville, Florida 32611, USA
Maryland 20742, USA 50. Torsten V. Zache
37. Niklas Mueller Institute for Theoretical Physics, University of
Department of Physics, Maryland Center for Innsbruck, 6020 Innsbruck, Austria
Fundamental Physics, and Joint Quantum INstitute, Institute for Quantum Optics and Quantum Infor-
University of Maryland, College Park, Maryland mation of the Austrian Academy of Sciences, 6020
20742, USA Innsbruck, Austria
38. Yasser Omar 51. Erez Zohar
Portuguese Quantum Institute, and Instituto Racah Institute of Physics, The Hebrew Univer-
Superior Técnico, Universidade de Lisboa, P-1049- sity of Jerusalem, Jerusalem 91904, Givat Ram,
001 Lisbon, Portugal Israel
39. Vincent R. Pascuzzi 52. Peter Zoller
Brookhaven National Laboratory, Upton, New Institute for Theoretical Physics, University of
York 11973, USA Innsbruck, 6020 Innsbruck, Austria
40. Indrakshi Raychowdhury Institute for Quantum Optics and Quantum Infor-
Birla Institute of Technology and Science, Pilani, mation of the Austrian Academy of Sciences, 6020
K. K. Birla Goa Campus, Zuarinagar, Goa - Innsbruck, Austria
403726, India.
41. Enrique Rico
IKERBASQUE, Basque Foundation for Science,
Plaza Euskadi 5, 48009 Bilbao, Spain [1] S. Ritz, et al., Report of the Particle Physics Project
Department of Physical Chemistry, University Prioritization Panel, (2014).
of the Basque Country UPV/EHU, Apartado 644, [2] D. Geesaman, et al., The 2015 Long Range Plan for
48080 Bilbao, Spain Nuclear Science, (2015).
EHU Quantum Center, University of the Basque [3] David J. Gross and Frank Wilczek, Ultraviolet Behavior
Country, UPV/EHU, Barrio Sarriena s/n, 48940 of Non-Abelian Gauge Theories, Phys. Rev. Lett. 30, 1343
(1973).
Leioa, Biscay, Spain
[4] H. David Politzer, Reliable Perturbative Results for
42. Ananda Roy Strong Interactions?, Phys. Rev. Lett. 30, 1346 (1973).

027001-50
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

[5] Matthias Troyer and Uwe-Jens Wiese, Computational [22] Arnau Riera, Christian Gogolin, and Jens Eisert, Thermal-
Complexity and Fundamental Limitations to Fermionic ization in Nature and on a Quantum Computer, Phys. Rev.
Quantum Monte Carlo Simulations, Phys. Rev. Lett. 94, Lett. 108, 080402 (2012).
170201 (2005). [23] Dmitri Kharzeev and Kirill Tuchin, From color glass con-
[6] Andrei Alexandru, Gökçe Başar, Paulo F. Bedaque, Sohan densate to quark–gluon plasma through the event horizon,
Vartak, and Neill C. Warrington, Monte Carlo Study of Nucl. Phys. A 753, 316 (2005).
Real Time Dynamics on the Lattice, Phys. Rev. Lett. 117, [24] Dmitri E. Kharzeev and Eugene M. Levin, Deep inelastic
081602 (2016). scattering as a probe of entanglement, Phys. Rev. D 95,
[7] Particle Data Group, P. [Link] Zyla, R. M. Barnett, J. 114008 (2017).
Beringer, O. Dahl, D. A. Dwyer, D. E. Groom, C.-J. Lin, [25] O. K. Baker and D. E. Kharzeev, Thermal radiation and
K. S. Lugovsky, and E. Pianori, et al., Review of particle entanglement in proton-proton collisions at energies avail-
physics, Prog. Theor. Exp. Phys. 2020, 083C01 (2020). able at the CERN large hadron collider, Phys. Rev. D 98,
[8] J. M. Campbell, M. Diefenthaler, T. J. Hobbs, S. Höche, J. 054007 (2018).
Isaacson, F. Kling, S. Mrenna, J. Reuter, S. Alioli, and J. R. [26] Jürgen Berges, Michal P. Heller, Aleksas Mazeliauskas,
Andersen, et al., Event generators for high-energy physics and Raju Venugopalan, QCD thermalization: Ab initio
experiments, (2022), arXiv preprint ArXiv:2203.11110. approaches and interdisciplinary connections, Rev. Mod.
[9] Thomas Gehrmann and Bogdan Malaescu, Precision Phys. 93, 035003 (2021).
QCD physics at the LHC, (2021), arXiv preprint [27] Hui Li, F. Duncan, and M. Haldane, Entanglement Spec-
ArXiv:2111.02319. trum as a Generalization of Entanglement Entropy: Iden-
[10] Torbjörn Sjöstrand, Stephen Mrenna, and Peter Skands, tification of Topological Order in Non-Abelian Fractional
Pythia 6.4 physics and manual, J. High Energy Phys. 2006, Quantum Hall Effect States, Phys. Rev. Lett. 101, 010504
026 (2006). (2008).
[11] Gennaro Corcella, Ian G. Knowles, Giuseppe March- [28] Scott D. Geraedts, Rahul Nandkishore, and Nicolas
esini, Stefano Moretti, Kosuke Odagiri, Peter Richardson, Regnault, Many-body localization and thermalization:
Michael H. Seymour, and Bryan R. Webber, Herwig 6: Insights from the entanglement spectrum, Phys. Rev. B 93,
An event generator for hadron emission reactions with 174202 (2016).
interfering gluons (including supersymmetric processes), [29] Adam M. Kaufman, M. Eric Tai, Alexander Lukin,
J. High Energy Phys. 2001, 010 (2001). Matthew Rispoli, Robert Schittko, Philipp M. Preiss, and
[12] Tanju Gleisberg, Stefan Höche, F. Krauss, M. Schönherr, Markus Greiner, Quantum thermalization through entan-
S. Schumann, F. Siegert, and J. Winter, Event generation glement in an isolated many-body system, Science 353,
with sherpa 1.1, J. High Energy Phys. 2009, 007 (2009). 794 (2016).
[13] Stephen P. Jordan, Keith S. M. Lee, and John Preskill, [30] Zhi-Cheng Yang, Claudio Chamon, Alioscia Hamma, and
Quantum algorithms for quantum field theories, Science Eduardo R. Mucciolo, Two-Component Structure in the
336, 1130 (2012). Entanglement Spectrum of Highly Excited States, Phys.
[14] Stephen P. Jordan, Keith S. M. Lee, and John Preskill, Rev. Lett. 115, 267206 (2015).
Quantum computation of scattering in scalar quantum [31] Niklas Mueller, Torsten V. Zache, and Robert Ott, Ther-
field theories, (2011), arXiv preprint ArXiv:1112.4833. malization of gauge theories from their entanglement
[15] Christian W. Bauer, Sean Fleming, and Michael Luke, spectrum, (2021), ArXiv:2107.11416.
Summing Sudakov logarithms in b → xsγ in effective [32] Yong-Zhong Qian, George M. Fuller, Grant J. Mathews,
field theory, Phys. Rev. D 63, 014006 (2000). Ron W. Mayle, James R. Wilson, and S. E. Woosley, Con-
[16] Christian W. Bauer, Sean Fleming, Dan Pirjol, and Iain nection Between Flavor-Mixing of Cosmologically Sig-
W. Stewart, An effective field theory for collinear and soft nificant Neutrinos and Heavy Element Nucleosynthesis in
gluons: Heavy to light decays, Phys. Rev. D 63, 114020 Supernovae, Phys. Rev. Lett. 71, 1965 (1993).
(2001). [33] Yong-Zhong Qian and George M. Fuller, Neutrino-
[17] Christian W. Bauer and Iain W. Stewart, Invariant oper- neutrino scattering and matter-enhanced neutrino flavor
ators in collinear effective theory, Phys. Lett. B 516, 134 transformation in supernovae, Phys. Rev. D 51, 1479
(2001). (1995).
[18] Christian W. Bauer, Dan Pirjol, and Iain W. Stewart, Soft- [34] Gianluigi Fogli, Eligio Lisi, Antonio Marrone, and
collinear factorization in effective field theory, Phys. Rev. Alessandro Mirizzi, Collective neutrino flavor transitions
D 65, 054022 (2002). in supernovae and the role of trajectory averaging, J.
[19] Kristan Temme, Tobias J. Osborne, Karl G. Vollbrecht, Cosmol. Astropart. Phys. 2007, 010 (2007).
David Poulin, and Frank Verstraete, Quantum metropolis [35] Zohreh Davoudi, Mohammad Hafezi, Christopher Mon-
sampling, Nature 471, 87 (2011). roe, Guido Pagano, Alireza Seif, and Andrew Shaw,
[20] Ronak Desai, Yuan Feng, Mohammad Hassan, Abhishek Towards analog quantum simulations of lattice gauge
Kodumagulla, and Michael McGuigan, Z3 gauge the- theories with trapped ions, Phys. Rev. Res. 2, 023015
ory coupled to fermions and quantum computing, (2021), (2020).
arXiv preprint ArXiv:2106.00549. [36] Alejandro Bermúdez, L Tagliacozzo, Germán Sierra, and
[21] Alexander M. Czajka, Zhong-Bo Kang, Henry Ma, and P. Richerme, Long-range Heisenberg models in quasiperi-
Fanyi Zhao, Quantum simulation of chiral phase transi- odically driven crystals of trapped ions, Phys. Rev. B 95,
tions, (2021), arXiv preprint ArXiv:2112.03944. 024431 (2017).

027001-51
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[37] Alessandro Roggero and Joseph Carlson, Dynamic linear [54] Ashley Milsted, Junyu Liu, John Preskill, and Guifre
response quantum algorithm, Phys. Rev. C. Nucl. Phys. Vidal, Collisions of false-vacuum bubble walls in a quan-
100, 034610 (2019). tum spin chain, (2020), arXiv preprint ArXiv:2012.07243.
[38] Alessandro Roggero, Andy C. Y. Li, Joseph Carlson, [55] Clemens Neuenhahn and Florian Marquardt, Quantum
Rajan Gupta, and Gabriel N. Perdue, Quantum computing simulation of expanding space–time with tunnel-coupled
for neutrino-nucleus scattering, Phys. Rev. D 101, 074038 condensates, New J. Phys. 17, 125007 (2015).
(2020). [56] Pablo Arrighi, Giuseppe Di Molfetta, and Stefano Fac-
[39] Eugene F. Dumitrescu, Alex J. McCaskey, Gaute Hagen, chini, Quantum walking in curved spacetime: Discrete
Gustav R. Jansen, Titus D. Morris, T. Papenbrock, metric, Quantum 2, 84 (2018).
Raphael C. Pooser, David Jarvis Dean, and Pavel Lougov- [57] Davide Rossini and Ettore Vicari, Dynamics after
ski, Cloud Quantum Computing of an Atomic Nucleus, quenches in one-dimensional quantum Ising-like systems,
Phys. Rev. Lett. 120, 210501 (2018). Phys. Rev. B 102, 054444 (2020).
[40] Hsuan-Hao Lu, Natalie Klco, Joseph M. Lukens, Titus D. [58] Igor Boettcher, Przemyslaw Bienias, Ron Belyansky, Ali-
Morris, Aaina Bansal, Andreas Ekström, Gaute Hagen, cia J. Kollár, and Alexey V. Gorshkov, Quantum simu-
Thomas Papenbrock, Andrew M. Weiner, and Martin lation of hyperbolic space with circuit quantum electro-
J. Savage, et al., Simulations of subatomic many-body dynamics: From graphs to geometry, Phys. Rev. A 102,
physics on a quantum frequency processor, Phys. Rev. A 032208 (2020).
100, 012320 (2019). [59] Kai Xu, Zheng-Hang Sun, Wuxin Liu, Yu-Ran Zhang,
[41] I. Stetcu, A. Baroni, and J. Carlson, Variational Hekang Li, Hang Dong, Wenhui Ren, Pengfei Zhang,
approaches to constructing the many-body nuclear ground Franco Nori, Dongning Zheng, Heng Fan, and H. Wang,
state for quantum computing, (2021), arXiv preprint Probing the dynamical phase transition with a supercon-
ArXiv:2110.06098. ducting quantum simulator, (2019), ArXiv:1912.05150.
[42] Alessandro Baroni, Joseph Carlson, Rajan Gupta, Andy [60] Xinfang Nie, Bo-Bo Wei, Xi Chen, Ze Zhang, Xiuzhu
C. Y. Li, Gabriel N. Perdue, and Alessandro Roggero, Zhao, Chudan Qiu, Yu Tian, Yunlan Ji, Tao Xin, and
Nuclear two point correlation functions on a quantum- Dawei Lu, and Jun Li, Experimental observation of equi-
computer, (2021), arXiv preprint ArXiv:2111.02982. librium and dynamical quantum phase transitions via out-
[43] Lev Kofman, Andrei D. Linde, and Alexei A. Starobin- of-time-ordered correlators, (2019), ArXiv:1912.12038.
sky, Reheating After Inflation, Phys. Rev. Lett. 73, 3195 [61] Markus Heyl, Dynamical quantum phase transitions: A
(1994). brief survey, EPL (Europhys. Lett.) 125, 26001 (2019).
[44] Lev Kofman, Andrei Linde, and Alexei A. Starobinsky, [62] Evan Berkowitz, Michael I. Buchoff, and Enrico Rinaldi,
Towards the theory of reheating after inflation, Phys. Rev. Lattice QCD input for axion cosmology, Phys. Rev. D 92,
D 56, 3258 (1997). 034507 (2015).
[45] Patrick B. Greene, Lev Kofman, Andrei Linde, and Alexei [63] Olivier Wantz and E. P. S. Shellard, Axion cosmology
A. Starobinsky, Structure of resonance in preheating after revisited, Phys. Rev. D 82, 123508 (2010).
inflation, Phys. Rev. D 56, 6175 (1997). [64] Michael S. Turner, Cosmic and local mass density of
[46] Aleksandr Chatrchyan, Kevin T. Geier, Markus K. “invisible” axions, Phys. Rev. D 33, 889 (1986).
Oberthaler, Jürgen Berges, and Philipp Hauke, Analog [65] Andrei D. Sakharov, in In The Intermissions. Collected
cosmological reheating in an ultracold bose gas, Phys. Works on Research into the Essentials of Theoretical
Rev. A 104, 023302 (2021). Physics in Russian Federal Nuclear Center, Arzamas-16
[47] Junyu Liu and Yue-Zhou Li, Quantum Simulation of (World Scientific, 1998), p. 84.
Cosmic Inflation, Phys. Rev. D 104, 086013 (2021). [66] Carroll L. Wainwright, Cosmotransitions: Computing cos-
[48] Salvatore Butera and Iacopo Carusotto, Particle creation mological phase transition temperatures and bubble pro-
in the spin modes of a dynamically oscillating two- files with multiple fields, Comput. Phys. Commun. 183,
component Bose-Einstein condensate, Phys. Rev. D 104, 2006 (2012).
083503 (2021). [67] Anatoli Polkovnikov and Vladimir Gritsev, Breakdown of
[49] João Barata, Niklas Mueller, Andrey Tarasov, and Raju the adiabatic limit in low-dimensional gapless systems,
Venugopalan, Single-particle digitization strategy for Nat. Phys. 4, 477 (2008).
quantum computation of a φ 4 scalar field theory, Phys. [68] Sebastian Baum, Marcela Carena, Nausheen R. Shah, Car-
Rev. A 103, 042410 (2021). los E. M. Wagner, and Yikun Wang, Nucleation is more
[50] N. C. Tsamis and R. P. Woodard, The quantum gravita- than critical: A case study of the electroweak phase tran-
tional back-reaction on inflation, Ann. Phys. (N. Y) 253, 1 sition in the NMSSM, J. High Energy Phys. 2021, 1
(1997). (2021).
[51] A. D. Sakharov, Violation of CP Invariance, C asymmetry, [69] Alexander Vilenkin, E. Paul, and S. Shellard, Cosmic
and baryon asymmetry of the universe, Sov. Phys. Usp. 34, Strings and Other Topological Defects (Cambridge Uni-
392 (1991). versity Press, 1994).
[52] Gianluca Lagnese, Federica Maria Surace, Márton Kor- [70] Ya. B. Zel’dovich and I. D. Novikov, The hypothesis of
mos, and Pasquale Calabrese, False vacuum decay in cores retarded during expansion and the hot cosmological
quantum spin chains, Phys. Rev. B 104, L201106 (2021). model, Sov. Astron. 10, 602 (1967).
[53] Dalila Pîrvu, Jonathan Braden, and Matthew C. John- [71] Juan Maldacena, The large-n limit of superconformal field
son, Bubble clustering in cosmological first order phase theories and supergravity, Int. J. Theor. Phys. 38, 1113
transitions, Phys. Rev. D 105, 043510 (2022). (1999).

027001-52
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

[72] Shinsei Ryu and Tadashi Takayanagi, Holographic [92] Edward Witten, Bound states of strings and p-branes,
Derivation of Entanglement Entropy from the Anti-de Nucl. Phys. B 460, 335 (1996).
Sitter Space/Conformal Field Theory Correspondence, [93] Masanori Hanada, Bulk geometry in gauge/gravity duality
Phys. Rev. Lett. 96, 181602 (2006). and color degrees of freedom, Phys. Rev. D 103, 106007
[73] Ahmed Almheiri, Xi Dong, and Daniel Harlow, Bulk (2021).
locality and quantum error correction in AdS/CFT, J. High [94] Evan Berkowitz, Enrico Rinaldi, Masanori Hanada, Goro
Energy Phys. 2015, 1 (2015). Ishiki, Shinji Shimasaki, and Pavlos Vranas, Monte Carlo
[74] Fernando Pastawski, Beni Yoshida, Daniel Harlow, String/M-Theory Collaboration, et al., Precision lattice
and John Preskill, Holographic quantum error-correcting test of the gauge/gravity duality at large n, Phys. Rev. D
codes: Toy models for the bulk/boundary correspondence, 94, 094501 (2016).
J. High Energy Phys. 2015, 1 (2015). [95] Georg Bergner, Norbert Bodendorfer, Masanori Hanada,
[75] Daniel Harlow and Hirosi Ooguri, Constraints on Sym- Stratos Pateloudis, Enrico Rinaldi, Andreas Schäfer,
metries from Holography, Phys. Rev. Lett. 122, 191601 Pavlos Vranas, and Hiromasa Watanabe, Confine-
(2019). ment/deconfinement transition in the d0-brane matrix
[76] Patrick Hayden and John Preskill, Black holes as mir- model–a signature of m-theory? (2021), arXiv preprint
rors: Quantum information in random subsystems, J. High ArXiv:2110.01312.
Energy Phys. 2007, 120 (2007). [96] Y. Aoki, T. Blum, G. Colangelo, S. Collins, M. Della
[77] Yasuhiro Sekino and Leonard Susskind, Fast scramblers, Morte, P. Dimopoulos, S. Dürr, X. Feng, H. Fukaya, and
J. High Energy Phys. 2008, 065 (2008). M. Golterman, et al., Flag review 2021, (2021), arXiv
[78] Juan Maldacena, Stephen H. Shenker, and Douglas Stan- preprint ArXiv:2111.09849.
ford, A bound on chaos, J. High Energy Phys. 2016, 1 [97] Christoph Lehner, Stefan Meinel, Tom Blum, Norman
(2016). H. Christ, Aida X. El-Khadra, Maxwell T. Hansen,
[79] Leonard Susskind, Computational complexity and black Andreas S. Kronfeld, Jack Laiho, Ethan T. Neil, and
hole horizons, Fortschritte der Phys. 64, 24 (2016). Stephen R. Sharpe, et al., Opportunities for lattice QCD
[80] Douglas Stanford and Leonard Susskind, Complexity and in quark and lepton flavor physics, Eur. Phys. J. A 55, 1
shock wave geometries, Phys. Rev. D 90, 126007 (2014). (2019).
[81] Adam R. Brown, Daniel A. Roberts, Leonard Susskind, [98] Vincenzo Cirigliano, Zohreh Davoudi, Tanmoy Bhat-
Brian Swingle, and Ying Zhao, Holographic Complex- tacharya, Taku Izubuchi, Phiala E. Shanahan, Sergey
ity Equals Bulk Action?, Phys. Rev. Lett. 116, 191301 Syritsyn, and Michael L. Wagman, The role of lattice
(2016). QCD in searches for violations of fundamental symme-
[82] Subir Sachdev and Jinwu Ye, Gapless Spin-Fluid Ground tries and signals for new physics, Eur. Phys. J. A 55, 1
State in a Random Quantum Heisenberg Magnet, Phys. (2019).
Rev. Lett. 70, 3339 (1993). [99] William Detmold, Robert G. Edwards, Jozef J. Dudek,
[83] Alexei Kitaev, A Simple Model of Quantum Holography Michael Engelhardt, Huey-Wen Lin, Stefan Meinel,
- KITP talks, (2015). Kostas Orginos, and Phiala Shanahan, Hadrons and nuclei,
[84] Netta Engelhardt and Aron C. Wall, Quantum extremal Eur. Phys. J. A 55, 1 (2019).
surfaces: Holographic entanglement entropy beyond the [100] Zohreh Davoudi, William Detmold, Phiala Shanahan,
classical regime, J. High Energy Phys. 2015, 1 (2015). Kostas Orginos, Assumpta Parreno, Martin J. Savage, and
[85] Geoffrey Penington, Entanglement wedge reconstruction Michael L. Wagman, Nuclear matrix elements from lat-
and the information paradox, J. High Energy Phys. 2020, tice QCD for electroweak and beyond-Standard-Model
1 (2020). processes, Phys. Rep. 900, 1 (2021).
[86] Ahmed Almheiri, Netta Engelhardt, Donald Marolf, and [101] Alexei Bazavov, Frithjof Karsch, Swagato Mukherjee, and
Henry Maxfield, The entropy of bulk quantum fields and Peter Petreczky, Hot-dense lattice QCD, Eur. Phys. J. A
the entanglement wedge of an evaporating black hole, J. 55, 1 (2019).
High Energy Phys. 2019, 1 (2019). [102] Heng-Tong Ding, Frithjof Karsch, and Swagato Mukher-
[87] Geoff Penington, Stephen H. Shenker, Douglas Stanford, jee, Thermodynamics of strong-interaction matter from
and Zhenbin Yang, Replica wormholes and the black hole lattice QCD, Int. J. Mod. Phys. E 24, 1530007
interior, (2019), arXiv preprint ArXiv:1911.11977. (2015).
[88] Ahmed Almheiri, Thomas Hartman, Juan Maldacena, [103] C. E. DeTar and U. M. Heller, QCD thermodynamics from
Edgar Shaghoulian, and Amirhossein Tajdini, The entropy the lattice, Eur. Phys. J. A 41, 405 (2009).
of hawking radiation, Rev. Mod. Phys. 93, 035002 (2021). [104] Martin Lüscher, Volume dependence of the energy spec-
[89] Tom Banks, Willy Fischler, Steven H. Shenker, and trum in massive quantum field theories, Commun. Math.
Leonard Susskind, M theory as a matrix model: A con- Phys. 105, 153 (1986).
jecture, Phys. Rev. D 55, 5112 (1997). [105] Martin Lüscher, Two-particle states on a torus and their
[90] Nissan Itzhaki, Juan M. Maldacena, Jacob Sonnenschein, relation to the scattering matrix, Nucl. Phys. B 354, 531
and Shimon Yankielowicz, Supergravity and the large n (1991).
limit of theories with sixteen supercharges, Phys. Rev. D [106] John Bulava, Raúl Briceño, William Detmold, Michael
58, 046004 (1998). Döring, Robert G Edwards, Anthony Francis, Francesco
[91] David Berenstein, Juan Maldacena, and Horatiu Nastase, Knechtli, Randy Lewis, Sasa Prelovsek, and Sinéad M
Strings in flat space and pp waves from n = 4 super yang Ryan, et al., Hadron spectroscopy with lattice QCD,
mills, J. High Energy Phys. 2002, 013 (2002). (2022), arXiv preprint ArXiv:2203.03230.

027001-53
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[107] Raúl A. Briceno, Jozef J. Dudek, and Ross D. Young, [124] Erez Zohar and Michele Burrello, Formulation of lattice
Scattering processes and resonances from lattice QCD, gauge theories for quantum simulations, Phys. Rev. D 91,
Rev. Mod. Phys. 90, 025001 (2018). 054506 (2015).
[108] Zohreh Davoudi, The path from finite to infinite volume: [125] Henry Lamm, Scott Lawrence, and Yukari Yamauchi,
Hadronic observables from lattice QCD, (2018), arXiv NuQS Collaboration, et al., General methods for digital
preprint ArXiv:1812.11899. quantum simulation of gauge theories, Phys. Rev. D 100,
[109] Maxwell T. Hansen and Stephen R. Sharpe, Lattice QCD 034518 (2019).
and three-particle decays of resonances, Annu. Rev. Nucl. [126] David B. Kaplan and Jesse R. Stryker, Gauss’s law, dual-
Part. Sci. 69, 65 (2019). ity, and the Hamiltonian formulation of U(1) lattice gauge
[110] Mari Carmen Bañuls and Krzysztof Cichy, Review on theory, Phys. Rev. D 102, 094515 (2020).
novel methods for lattice gauge theories, Rep. Prog. Phys. [127] Jan F. Haase, Luca Dellantonio, Alessio Celi, Danny Paul-
83, 024401 (2020). son, Angus Kan, Karl Jansen, and Christine A. Muschik, A
[111] Mari Carmen Banuls, Rainer Blatt, Jacopo Catani, Alessio resource efficient approach for quantum and classical sim-
Celi, Juan Ignacio Cirac, Marcello Dalmonte, Leonardo ulations of gauge theories in particle physics, Quantum 5,
Fallani, Karl Jansen, Maciej Lewenstein, and Simone 393 (2021).
Montangero, et al., Simulating lattice gauge theories [128] Christian W. Bauer and Dorota M. Grabowska, Efficient
within quantum technologies, Eur. Phys. J. D 74, 1 representation for simulating U(1) gauge theories on dig-
(2020). ital quantum computers at all values of the coupling,
[112] Giuseppe Magnifico, Timo Felser, Pietro Silvi, and (2021), arXiv preprint ArXiv:2111.08015.
Simone Montangero, Lattice quantum electrodynamics in [129] Manu Mathur and T. P. Sreeraj, Lattice gauge theories and
(3+1)-dimensions at finite density with tensor networks, spin models, Phys. Rev. D 94, 085029 (2016).
Nat. Commun. 12, 1 (2021). [130] Yuzhi Liu, Y. Meurice, M. P. Qin, J. Unmuth-Yockey, T.
[113] Timo Felser, Pietro Silvi, Mario Collura, and Simone Xiang, Z. Y. Xie, J. F. Yu, and Haiyuan Zou, Exact block-
Montangero, Two-Dimensional Quantum-Link Lattice ing formulas for spin and gauge models, Phys. Rev. D 88,
Quantum Electrodynamics at Finite Density, Phys. Rev. 056005 (2013).
X 10, 041040 (2020). [131] Zhi-Yuan Xie, Jing Chen, Ming-Pu Qin, Jing W Zhu,
[114] John Kogut and Leonard Susskind, Hamiltonian formula- Li-Ping Yang, and Tao Xiang, Coarse-graining renor-
tion of Wilson’s lattice gauge theories, Phys. Rev. D 11, malization by higher-order singular value decomposition,
395 (1975). Phys. Rev. B 86, 045139 (2012).
[115] Natalie Klco, Eugene F. Dumitrescu, Alex J. McCaskey, [132] Yannick Meurice, Ryo Sakai, and Judah Unmuth-
Titus D. Morris, Raphael C. Pooser, Mikel Sanz, Enrique Yockey, Tensor lattice field theory for renormalization
Solano, Pavel Lougovski, and Martin J. Savage, Quantum- and quantum computing, Rev. Mod. Phys. 94, 025005
classical computation of Schwinger model dynamics (2022).
using quantum computers, Phys. Rev. A 98, 032331 [133] Michael Kreshchuk, Shaoyang Jia, William M. Kirby,
(2018). Gary Goldstein, James P. Vary, and Peter J. Love, Light-
[116] Natalie Klco, Martin J. Savage, and Jesse R. Stryker, front field theory on current quantum computers, Entropy
SU(2) non-Abelian gauge field theory in one dimension 23, 597 (2021).
on digital quantum computers, Phys. Rev. D 101, 074512 [134] Michael Kreshchuk, Shaoyang Jia, William M. Kirby,
(2020). Gary Goldstein, James P. Vary, and Peter J. Love, Simulat-
[117] Tim Byrnes and Yoshihisa Yamamoto, Simulating lattice ing hadronic physics on noisy intermediate-scale quantum
gauge theories on a quantum computer, Phys. Rev. A 73, devices using basis light-front quantization, Phys. Rev. A
022328 (2006). 103, 062601 (2021).
[118] Anthony Ciavarella, Natalie Klco, and Martin J. Savage, [135] Michael Kreshchuk, William M. Kirby, Gary Goldstein,
Trailhead for quantum simulation of SU(3) Yang-Mills Hugo Beauchemin, and Peter J. Love, Quantum simula-
lattice gauge theory in the local multiplet basis, Phys. Rev. tion of quantum field theory in the light-front formulation,
D 103, 094501 (2021). Phys. Rev. A 105, 032418 (2022).
[119] Manu Mathur, Harmonic oscillator pre-potentials in SU(2) [136] R. Brower, Shailesh Chandrasekharan, and U.-J. Wiese,
lattice gauge theory, J. Phys. A: Math. Gen. 38, 10015 QCD as a quantum link model, Phys. Rev. D 60, 094502
(2005). (1999).
[120] Ramesh Anishetty and Indrakshi Raychowdhury, SU(2) [137] R. Brower, S. Chandrasekharan, S. Riederer, and U.-
lattice gauge theory: Local dynamics on nonintersecting J. Wiese, D-theory: Field quantization by dimensional
electric flux loops, Phys. Rev. D 90, 114503 (2014). reduction of discrete variables, Nucl. Phys. B 693, 149
[121] Ramesh Anishetty, Manu Mathur, and Indrakshi Ray- (2004).
chowdhury, Irreducible SU(3) Schwinger Bosons, J. Math. [138] Alexander J. Buser, Tanmoy Bhattacharya, Lukasz Cincio,
Phys. 50, 053503 (2009). and Rajan Gupta, State preparation and measurement in a
[122] Manu Mathur, Indrakshi Raychowdhury, and Ramesh quantum simulation of the O(3) sigma model, Phys. Rev.
Anishetty, SU(N) irreducible Schwinger bosons, J. Math. D 102, 114514 (2020).
Phys. 51, 093504 (2010). [139] Tanmoy Bhattacharya, Alexander J. Buser, Shailesh
[123] Indrakshi Raychowdhury and Jesse R. Stryker, Loop, Chandrasekharan, Rajan Gupta, and Hersh Singh, Qubit
string, and hadron dynamics in SU(2) Hamiltonian lattice Regularization of Asymptotic Freedom, Phys. Rev. Lett.
gauge theories, Phys. Rev. D 101, 114502 (2020). 126, 172001 (2021).

027001-54
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

[140] Andrei Alexandru, Paulo F. Bedaque, Andrea Carosso, [157] Jad C. Halimeh, Haifeng Lang, and Philipp Hauke, Gauge
and Andy Sheng, Universality of a truncated sigma- protection in non-Abelian lattice gauge theories, New J.
model, (2021), arXiv preprint ArXiv:2109.07500. Phys. 24, 033015 (2022).
[141] M. Luscher, Some analytic results concerning the mass [158] Jad C. Halimeh, Lukas Homeier, Christian Schweizer,
spectrum of Yang-Mills gauge theories on a torus, Nucl. Monika Aidelsburger, Philipp Hauke, and Fabian Grusdt,
Phys. B 219, 233 (1983). Stabilizing lattice gauge theories through simplified local
[142] Pierre Van Baal, in At The Frontier of Particle Physics: pseudo generators, (2021), arXiv preprint ArXiv:2108.
Handbook of QCD (In 3 Volumes) (World Scientific, 02203.
2001), p. 683. [159] Maarten Van Damme, Julius Mildenberger, Fabian
[143] Tom Banks, Bartomeu Fiol, and Alexander Morisse, Grusdt, Philipp Hauke, and Jad C. Halimeh, Suppress-
Towards a quantum theory of de Sitter space, J. High ing nonperturbative gauge errors in the thermodynamic
Energy Phys. 2006, 004 (2006). limit using local pseudogenerators, (2021), arXiv preprint
[144] Dionysios Anninos and Beatrix Mühlmann, Notes on ArXiv:2110.08041.
matrix models (matrix musings), J. Stat. Mech.: Theory [160] K. Stannigel, P. Hauke, D. Marcos, M. Hafezi, S. Diehl,
Exp. 2020, 083109 (2020). M. Dalmonte, and P. Zoller, Constrained Dynamics via the
[145] Jad C. Halimeh, Valentin Kasper, and Philipp Hauke, Fate Zeno Effect in Quantum Simulation: Implementing Non-
of lattice gauge theories under decoherence, (2020), arXiv Abelian Lattice Gauge Theories with Cold Atoms, Phys.
preprint ArXiv:2009.07848. Rev. Lett. 112, 120406 (2014).
[146] Maarten Van Damme, Haifeng Lang, Philipp Hauke, [161] Valentin Kasper, Torsten V. Zache, Fred Jendrzejewski,
and Jad C. Halimeh, Reliability of lattice gauge theo- Maciej Lewenstein, and Erez Zohar, Non-Abelian gauge
ries in the thermodynamic limit, (2021), arXiv preprint invariance from dynamical decoupling, (2020), arXiv
ArXiv:2104.07040. preprint ArXiv:2012.08620.
[147] Jesse R. Stryker, Oracles for Gauss’s law on digital quan- [162] Nhung H. Nguyen, Minh C. Tran, Yingyue Zhu, Alaina
tum computers, Phys. Rev. A 99, 042301 (2019). M. Green, C. Huerta Alderete, Zohreh Davoudi, and
[148] Indrakshi Raychowdhury and Jesse R. Stryker, Solving Norbert M. Linke, Digital quantum simulation of the
Gauss’s law on digital quantum computers with loop- Schwinger model and symmetry protection with trapped
string-hadron digitization, Phys. Rev. Res. 2, 033039 ions, (2021), arXiv preprint ArXiv:2112.14262.
(2020). [163] Alexander F. Shaw, Pavel Lougovski, Jesse R. Stryker,
[149] Debasish Banerjee, M. Dalmonte, M. Müller, E. Rico, and Nathan Wiebe, Quantum algorithms for simulating the
P. Stebler, U.-J. Wiese, and P. Zoller, Atomic Quan- lattice Schwinger model, Quantum 4, 306 (2020).
tum Simulation of Dynamical Gauge Fields Coupled to [164] Angus Kan and Yunseong Nam, Lattice quantum chro-
Fermionic Matter: From String Breaking to Evolution modynamics and electrodynamics on a universal quantum
After a Quench, Phys. Rev. Lett. 109, 175302 (2012). computer, (2021), arXiv preprint ArXiv:2107.12769.
[150] Philipp Hauke, David Marcos, Marcello Dalmonte, and [165] Yuan Su, Hsin-Yuan Huang, and Earl T. Campbell, Nearly
Peter Zoller, Quantum Simulation of a Lattice Schwinger tight trotterization of interacting electrons, Quantum 5,
Model in a Chain of Trapped Ions, Phys. Rev. X 3, 041018 495 (2021).
(2013). [166] Burak Şahinoğlu and Rolando D. Somma, Hamiltonian
[151] Debasish Banerjee, Michael Bögli, Marcello Dalmonte, simulation in the low-energy subspace, npj Quantum Inf.
Enrique Rico, Pascal Stebler, U.-J. Wiese, and Peter 7, 1 (2021).
Zoller, Atomic Quantum Simulation of U(N) and SU(N) [167] Takuya Hatomura, State-dependent error bound for digi-
Non-Abelian Lattice Gauge Theories, Phys. Rev. Lett. tal quantum simulation of driven systems, (2022), arXiv
110, 125303 (2013). preprint ArXiv:2201.04835.
[152] David Marcos, Peter Rabl, Enrique Rico, and Peter Zoller, [168] Changhao Yi and Elizabeth Crosson, Spectral analysis of
Superconducting Circuits for Quantum Simulation of product formulas for quantum simulation, (2021), arXiv
Dynamical Gauge Fields, Phys. Rev. Lett. 111, 110504 preprint ArXiv:2102.12655.
(2013). [169] Qi Zhao, You Zhou, Alexander F. Shaw, Tongyang Li, and
[153] Abhishek Rajput, Alessandro Roggero, and Nathan Andrew M. Childs, Hamiltonian simulation with random
Wiebe, Hybridized methods for quantum simulation in inputs, (2021), arXiv preprint ArXiv:2111.04773.
the interaction picture, (2021), arXiv preprint ArXiv:2109. [170] Andrew M. Childs and Yuan Su, Nearly Optimal Lattice
03308. Simulation by Product Formulas, Phys. Rev. Lett. 123,
[154] Minh C. Tran, Yuan Su, Daniel Carney, and Jacob M. 050503 (2019).
Taylor, Faster Digital Quantum Simulation by Symmetry [171] Torin F. Stetina, Anthony Ciavarella, Xiaosong Li, and
Protection, PRX Quantum 2, 010323 (2021). Nathan Wiebe, Simulating effective QED on quantum
[155] Henry Lamm, Scott Lawrence, and Yukari Yamauchi, computers, (2020), arXiv preprint ArXiv:2101.00111.
Suppressing coherent gauge drift in quantum simulations, [172] William Donnelly, Decomposition of entanglement
(2020), arXiv preprint ArXiv:2005.12688. entropy in lattice gauge theory, Phys. Rev. D 85, 085004
[156] Jad C. Halimeh, Haifeng Lang, Julius Mildenberger, (2012).
Zhang Jiang, and Philipp Hauke, Gauge-Symmetry Pro- [173] Horacio Casini, Marina Huerta, and Jose Alejandro Ros-
tection using Single-Body Terms, PRX Quantum 2, abal, Remarks on entanglement entropy for gauge fields,
040311 (2021). Phys. Rev. D 89, 085012 (2014).

027001-55
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[174] Sinya Aoki, Takumi Iritani, Masahiro Nozaki, Tokiro [189] M. Sohaib Alam, Stuart Hadfield, Henry Lamm, and Andy
Numasawa, Noburo Shiba, and Hal Tasaki, On the C. Y. Li, Quantum simulation of dihedral gauge theories,
definition of entanglement entropy in lattice gauge theo- (2021), arXiv preprint ArXiv:2108.13305.
ries, J. High Energy Phys. 2015, 1 (2015). [190] Yasar Atas, Jinglei Zhang, Randy Lewis, Amin Jahan-
[175] Ronak M. Soni and Sandip P. Trivedi, Aspects of entan- pour, Jan F. Haase, and Christine A. Muschik, Su(2)
glement entropy for gauge theories, J. High Energy Phys. hadrons on a quantum computer, (2021), arXiv preprint
2016, 1 (2016). ArXiv:2102.08920.
[176] Alicia J. Kollár, Mattias Fitzpatrick, and Andrew A. [191] Anthony N. Ciavarella and Ivan A. Chernyshev, Prepa-
Houck, Hyperbolic lattices in circuit quantum electrody- ration of the SU(3) Lattice Yang-Mills vacuum with
namics, Nature 571, 45 (2019). variational quantum methods, (2021), arXiv preprint
[177] Avikar Periwal, Eric S. Cooper, Philipp Kunkel, Julian ArXiv:2112.09083.
F. Wienand, Emily J. Davis, and Monika Schleier-Smith, [192] Bin Xu and Wei Xue, 3+1 dimension Schwinger pair pro-
Programmable interactions and emergent geometry in an duction with quantum computers, (2021), arXiv preprint
array of atom clouds, Nature 600, 630 (2021). ArXiv:2112.06863.
[178] Brian Swingle, Gregory Bentsen, Monika Schleier- [193] Julius Mildenberger, Wojciech Mruczkiewicz, Jad C. Hal-
Smith, and Patrick Hayden, Measuring the scrambling imeh, Zhang Jiang, and Philipp Hauke, Probing confine-
of quantum information, Phys. Rev. A 94, 040302 ment in a Z2 lattice gauge theory on a quantum computer,
(2016). (2022), arXiv preprint ArXiv:2203.08905.
[179] Erez Zohar, J. Ignacio Cirac, and Benni Reznik, Sim- [194] Christian Schweizer, Fabian Grusdt, Moritz Berngru-
ulating Compact Quantum Electrodynamics with Ultra- ber, Luca Barbiero, Eugene Demler, Nathan Goldman,
cold Atoms: Probing Confinement and Nonperturbative Immanuel Bloch, and Monika Aidelsburger, Floquet
Effects, Phys. Rev. Lett. 109, 125302 (2012). approach to Z@ lattice gauge theories with ultracold
[180] Erez Zohar and Benni Reznik, Confinement and Lattice atoms in optical lattices, Nat. Phys. 15, 1168 (2019).
Quantum-Electrodynamic Electric Flux Tubes Simulated [195] Frederik Görg, Kilian Sandholzer, Joaquín Minguzzi,
with Ultracold Atoms, Phys. Rev. Lett. 107, 275301 Rémi Desbuquois, Michael Messer, and Tilman Esslinger,
(2011). Realization of density-dependent Peierls phases to engi-
[181] Erez Zohar, J. Ignacio Cirac, and Benni Reznik, Simulat- neer quantized gauge fields coupled to ultracold matter,
ing (2+1)-Dimensional Lattice QED with Dynamical Mat- Nat. Phys. 15, 1161 (2019).
ter using Ultracold Atoms, Phys. Rev. Lett. 110, 055302 [196] Alexander Mil, Torsten V. Zache, Apoorva Hegde, Andy
(2013). Xia, Rohit P. Bhatt, Markus K. Oberthaler, Philipp
[182] Luca Tagliacozzo, Alessio Celi, Alejandro Zamora, and Hauke, Jürgen Berges, and Fred Jendrzejewski, Real-
Maciej Lewenstein, Optical Abelian lattice gauge theo- izing a scalable building block of a U(1) gauge the-
ries, Ann. Phys. (N. Y) 330, 160 (2013). ory with cold atomic mixtures, (2019), arXiv preprint
[183] Robert Ott, Torsten V. Zache, Fred Jendrzejewski, and ArXiv:1909.07641.
Jürgen Berges, Scalable Cold-Atom Quantum Simulator [197] Bing Yang, Hui Sun, Robert Ott, Han-Yi Wang, Torsten V.
for Two-Dimensional QED, Phys. Rev. Lett. 127, 130504 Zache, Jad C. Halimeh, Zhen-Sheng Yuan, Philipp Hauke,
(2021). and Jian-Wei Pan, Observation of gauge invariance in
[184] Daniel González-Cuadra, Torsten V. Zache, Jose Car- a 71-site Bose–Hubbard quantum simulator, Nature 587,
rasco, Barbara Kraus, and Peter Zoller, Hardware effi- 392 (2020).
cient quantum simulation of non-Abelian gauge theories [198] Zhao-Yu Zhou, Guo-Xian Su, Jad C. Halimeh, Robert Ott,
with qudits on Rydberg platforms, (2022), arXiv preprint Hui Sun, Philipp Hauke, Bing Yang, Zhen-Sheng Yuan,
ArXiv:2203.15541. Jürgen Berges, and Jian-Wei Pan, Thermalization dynam-
[185] Kevin Marshall, Raphael Pooser, George Siopsis, and ics of a gauge theory on a quantum simulator, (2021),
Christian Weedbrook, Quantum simulation of quantum arXiv preprint ArXiv:2107.13563.
field theory using continuous variables, Phys. Rev. A 92, [199] Bárbara Andrade, Zohreh Davoudi, Tobias Graß, Moham-
063825 (2015). mad Hafezi, Guido Pagano, and Alireza Seif, Engineer-
[186] Kübra Yeter-Aydeniz and George Siopsis, Quantum com- ing an effective three-spin hamiltonian in trapped-ion
putation of scattering amplitudes in scalar quantum elec- systems for applications in quantum simulation, (2021),
trodynamics, Phys. Rev. D 97, 036004 (2018). ArXiv:2108.01022.
[187] Esteban A. Martinez, Christine A. Muschik, Philipp [200] Federica M. Surace, Paolo P. Mazza, Giuliano Giudici,
Schindler, Daniel Nigg, Alexander Erhard, Markus Heyl, Alessio Lerose, Andrea Gambassi, and Marcello Dal-
Philipp Hauke, Marcello Dalmonte, Thomas Monz, and monte, Lattice Gauge Theories and String Dynamics in
Peter Zoller, et al., Real-time dynamics of lattice gauge Rydberg Atom Quantum Simulators, Phys. Rev. X 10,
theories with a few-qubit quantum computer, Nature 534, 021041 (2020).
516 (2016). [201] Zohreh Davoudi, Norbert M. Linke, and Guido Pagano,
[188] Christian Kokail, Christine Maier, Rick van Bijnen, Tiff Toward simulating quantum field theories with controlled
Brydges, Manoj K. Joshi, Petar Jurcevic, Christine A. phonon-ion dynamics: A hybrid analog-digital approach,
Muschik, Pietro Silvi, Rainer Blatt, and Christian F. Roos, Phys. Rev. Res. 3, 043072 (2021).
et al., Self-verifying variational quantum simulation of [202] Jorge Casanova, Lucas Lamata, I. L. Egusquiza, Rene
lattice models, Nature 569, 355 (2019). Gerritsma, Christian F. Roos, Juan José García-Ripoll, and

027001-56
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

Enrique Solano, Quantum Simulation of Quantum Field [219] Niklas Mueller, Andrey Tarasov, and Raju Venugopalan,
Theories in Trapped Ions, Phys. Rev. Lett. 107, 260501 Deeply inelastic scattering structure functions on a hybrid
(2011). quantum computer, Phys. Rev. D 102, 016007 (2020).
[203] Erez Zohar, J. Ignacio Cirac, and Benni Reznik, Cold- [220] Wenyang Qian, Robert Basili, Soham Pal, Glenn Luecke,
Atom Quantum Simulator for SU(2) Yang-Mills Lattice and James P. Vary, Solving hadron structures with vari-
Gauge Theory, Phys. Rev. Lett. 110, 125304 (2013). ational quantum eigensolvers, (2021), arXiv preprint
[204] Raka Dasgupta and Indrakshi Raychowdhury, Cold-atom ArXiv:2112.01927.
quantum simulator for string and hadron dynamics in non- [221] J. S. Pedernales, R. Di Candia, I. L. Egusquiza, J.
Abelian lattice gauge theory, Phys. Rev. A 105, 023322 Casanova, and E. Solano, Efficient Quantum Algorithm
(2022). for Computing n-Time Correlation Functions, Phys. Rev.
[205] Erez Zohar, J. Ignacio Cirac, and Benni Reznik, Quantum Lett. 113, 020505 (2014).
simulations of gauge theories with ultracold atoms: Local [222] Christian W. Bauer, Benjamin Nachman, and Marat Freyt-
gauge invariance from angular-momentum conservation, sis, Simulating Collider Physics on Quantum Comput-
Phys. Rev. A 88, 023617 (2013). ers using Effective Field Theories, Phys. Rev. Lett. 127,
[206] Junyu Liu, Zimu Li, Han Zheng, Xiao Yuan, and Jinzhao 212001 (2021).
Sun, Towards a variational Jordan–Lee–Preskill quantum [223] Manuel Bähr, Stefan Gieseke, Martyn A. Gigg, David
algorithm, Machine Learning: Sci. Technol. 3, 045030 Grellscheid, Keith Hamilton, Oluseyi Latunde-Dada,
(2022). Simon Plätzer, Peter Richardson, Michael H. Seymour,
[207] Siddhartha Harmalkar, Henry Lamm, and Scott Lawrence, andAlexander Sherstnev, et al., Herwig++ physics and
Quantum simulation of field theories without state prepa- manual, Eur. Phys. J. C 58, 639 (2008).
ration, (2020), arXiv preprint ArXiv:2001.11490. [224] Christian W. Bauer and Matthew D. Schwartz, Event
[208] Erik J. Gustafson and Henry Lamm, Toward quantum generation from effective field theory, Phys. Rev. D 76,
simulations of z2 gauge theory without state preparation, 074004 (2007).
Phys. Rev. D 103, 054507 (2021). [225] Benjamin Nachman, Davide Provasoli, Wibe A. de Jong,
[209] Chia Cheng Chang, Arjun Gambhir, Travis S. Humble, and Christian W. Bauer, Quantum Algorithm for High
and Shigetoshi Sota, Quantum annealing for systems of Energy Physics Simulations, Phys. Rev. Lett. 126, 062001
polynomial equations, Sci. Rep. 9, 1 (2019). (2021).
[210] A. Avkhadiev, P. E. Shanahan, and R. D. Young, Accel- [226] Jonathan Engel and Javier Menéndez, Status and future
erating Lattice Quantum Field Theory Calculations via of nuclear matrix elements for neutrinoless double-beta
Interpolator Optimization using Noisy Intermediate-Scale decay: A review, Rep. Prog. Phys. 80, 046301 (2017).
Quantum Computing, Phys. Rev. Lett. 124, 080501 [227] L. Baudis, G. Kessler, P. Klos, R. F. Lang, J. Menéndez,
(2020). S. Reichard, and A. Schwenk, Signatures of dark matter
[211] U. Aglietti, Marco Ciuchini, G. Corbo, E. Franco, G. scattering inelastically off nuclei, Phys. Rev. D 88, 115014
Martinelli, and Luca Silvestrini, Model independent deter- (2013).
mination of the shape function for inclusive b decays and [228] Lorenzo Andreoli, Vincenzo Cirigliano, Stefano Gandolfi,
of the structure functions in dis, Phys. Lett. B 432, 411 and Francesco Pederiva, Quantum Monte Carlo calcula-
(1998). tions of dark matter scattering off light nuclei, Phys. Rev.
[212] Keh-Fei Liu, Parton degrees of freedom from the path- C. Nucl. Phys. 99, 025501 (2019).
integral formalism, Phys. Rev. D 62, 074501 (2000). [229] Omar Shehab, K. Landsman, Yunseong Nam, Daiwei
[213] Xiangdong Ji, Parton Physics on a Euclidean Lattice, Zhu, Norbert M. Linke, M. Keesan, Raphael C. Pooser,
Phys. Rev. Lett. 110, 262002 (2013). and C. Monroe, Toward convergence of effective-field-
[214] A. J. Chambers, R. Horsley, Y. Nakamura, H. Perlt, P. E. theory simulations on digital quantum computers, Phys.
L. Rakow, G. Schierholz, A. Schiller, K. Somfleth, R. D. Rev. A 100, 062319 (2019).
Young, and J. M. Zanotti, et al., Nucleon Structure Func- [230] Alessandro Roggero, Chenyi Gu, Alessandro Baroni, and
tions from Operator Product Expansion on the Lattice, Thomas Papenbrock, Preparation of excited states for
Phys. Rev. Lett. 118, 242001 (2017). nuclear dynamics on a quantum computer, Phys. Rev. C.
[215] Yan-Qing Ma and Jian-Wei Qiu, Extracting Parton dis- Nucl. Phys. 102, 064624 (2020).
tribution functions from lattice QCD calculations, Phys. [231] G. F. Burgio, H.-J. Schulze, I. Vidaña, and J.-B. Wei, Neu-
Rev. D 98, 074021 (2018). tron stars and the nuclear equation of state, Prog. Part.
[216] Henry Lamm, Scott Lawrence, and Yukari Yamauchi, Nucl. Phys. 120, 103879 (2021).
NuQS Collaboration, et al., Parton physics on a quantum [232] Thomas Krüger, I. Tews, K. Hebeler, and A. Schwenk,
computer, Phys. Rev. Res. 2, 013272 (2020). Neutron matter from chiral effective field theory interac-
[217] M. G. Echevarria, I. L. Egusquiza, E. Rico, and G. tions, Phys. Rev. C. Nucl. Phys. 88, 025802 (2013).
Schnell, Quantum simulation of light-front Parton corre- [233] Stefano Gandolfi, Alexandros Gezerlis, and Joe Carlson,
lators, Phys. Rev. D 104, 014512 (2021). Neutron matter from low to high density, Annu. Rev. Nucl.
[218] Tianyin Li, Xingyu Guo, Wai Kin Lai, Xiaohui Liu, Enke Part. Sci. 65, 303 (2015).
Wang, Hongxi Xing, Dan-Bo Zhang, and Shi-Liang Zhu, [234] Diego Lonardoni, I. Tews, S. Gandolfi, and J. Carlson,
Partonic structure by quantum computing, (2021), arXiv Nuclear and neutron-star matter from local chiral interac-
preprint ArXiv:2106.03865. tions, Phys. Rev. Res. 2, 022033 (2020).

027001-57
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[235] I. Tews, J. Margueron, and S. Reddy, Confronting [251] C. Monroe, W. C. Campbell, L.-M. Duan, Z.-X. Gong,
gravitational-wave observations with modern nuclear A. V. Gorshkov, P. W. Hess, R. Islam, K. Kim, N. M.
physics constraints, Eur. Phys. J. A 55, 1 (2019). Linke, G. Pagano, P. Richerme, C. Senko, and N. Y. Yao,
[236] Isaac Vidaña, Hyperons: The strange ingredients of the Programmable quantum simulations of spin systems with
nuclear equation of state, Proc. R. Soc. A: Math., Phys. trapped ions, Rev. Mod. Phys. 93, 025001 (2021).
Eng. Sci. 474, 20180145 (2018). [252] A. Bermudez, D. Porras, and M. A. Martin-Delgado,
[237] Laura Tolos and Laura Fabbietti, Strangeness in nuclei and Competing many-body interactions in systems of trapped
neutron stars, Prog. Part. Nucl. Phys. 112, 103770 (2020). ions, Phys. Rev. A 79, 060303 (2009).
[238] Nir Barnea, L. Contessi, D. Gazit, F. Pederiva, and Ubira- [253] Or Katz, Marko Cetina, and Christopher Monroe, n-
jara Van Kolck, Effective Field Theory for Lattice Nuclei, body interactions between trapped ion qubits via spin-
Phys. Rev. Lett. 114, 052501 (2015). dependent squeezing, (2022), arXiv preprint ArXiv:2202.
[239] L. Contessi, A. Lovato, F. Pederiva, A. Roggero, J. 04230.
Kirscher, and U. van Kolck, Ground-state properties of [254] Pei Jiang Low, Brendan M. White, Andrew A. Cox,
4He and 16O extrapolated from lattice QCD with pionless Matthew L. Day, and Crystal Senko, Practical trapped-ion
EFT, Phys. Lett. B 772, 839 (2017). protocols for universal qudit-based quantum computing,
[240] Aaina Bansal, Sven Binder, Andreas Ekström, Gaute Phys. Rev. Res. 2, 033128 (2020).
Hagen, Gustav R. Jansen, and Thomas Papenbrock, Pion- [255] C. Senko, P. Richerme, J. Smith, A. Lee, I. Cohen, A. Ret-
less effective field theory for atomic nuclei and lattice zker, and C. Monroe, Realization of a Quantum Integer-
nuclei, Phys. Rev. C. Nucl. Phys. 98, 054301 (2018). Spin Chain with Controllable Interactions, Phys. Rev. X
[241] Mike H. Anderson, Jason R. Ensher, Michael R. 5, 021026 (2015).
Matthews, Carl E. Wieman, and Eric A. Cornell, Obser- [256] Dayou Yang, Gouri Shankar Giri, Michael Johan-
vation of Bose-Einstein condensation in a dilute atomic ning, Christof Wunderlich, Peter Zoller, and Philipp
vapor, Science 269, 198 (1995). Hauke, Analog quantum simulation of 1+1 lattice QED
[242] Brian DeMarco and Deborah S. Jin, Onset of Fermi with trapped ions, arXiv preprint 94, 052321 (2016),
degeneracy in a trapped atomic gas, Science 285, 1703 ArXiv:1604.03124.
(1999). [257] Ehud Altman, Kenneth R. Brown, Giuseppe Carleo,
[243] Wolfgang Ketterle and Martin W. Zwierlein, Making, Lincoln D. Carr, Eugene Demler, Cheng Chin, Brian
probing and understanding ultracold Fermi gases, La Riv- DeMarco, Sophia E Economou, Mark A. Eriksson, and
ista del Nuovo Cimento 31, 247 (2008). Kai-Mei C. Fu, et al., Quantum Simulators: Architectures
[244] Z. Hadzibabic, C. A. Stan, K. Dieckmann, S. Gupta, M. and Opportunities, PRX Quantum 2, 017003 (2021).
W. Zwierlein, A. Görlitz, and W. Ketterle, Two-Species [258] Ye Wang, Mark Um, Junhua Zhang, Shuoming An, Ming
Mixture of Quantum Degenerate Bose and Fermi Gases, Lyu, Jing-Ning Zhang, L. M. Duan, Dahyun Yum, and
Phys. Rev. Lett. 88, 160401 (2002). Kihwan Kim, Single-qubit quantum memory exceed-
[245] Waseem S. Bakr, Jonathon I. Gillen, Amy Peng, Simon ing ten-minute coherence time, Nat. Photon. 11, 646
Fölling, and Markus Greiner, A quantum gas microscope (2017).
for detecting single atoms in a Hubbard-regime optical [259] T. P. Harty, D. T. C. Allcock, C. J. Ballance, L. Guidoni, H.
lattice, Nature 462, 74 (2009). A. Janacek, N. M. Linke, D. N. Stacey, and D. M. Lucas,
[246] Jacob F. Sherson, Christof Weitenberg, Manuel Endres, High-Fidelity Preparation, Gates, Memory, and Readout
Marc Cheneau, Immanuel Bloch, and Stefan Kuhr, Single- of a Trapped-Ion Quantum Bit, Phys. Rev. Lett. 113,
atom-resolved fluorescence imaging of an atomic Mott 220501 (2014).
insulator, Nature 467, 68 (2010). [260] J. M. Pino, J. M. Dreiling, C. Figgatt, J. P. Gaebler, S. A.
[247] Lawrence W. Cheuk, Matthew A. Nichols, Melih Okan, Moses, M. S. Allman, C. H. Baldwin, M. Foss-Feig, D.
Thomas Gersdorf, Vinay V. Ramasesh, Waseem S. Bakr, Hayes, K. Mayer, C. Ryan-Anderson, and B. Neyenhuis,
Thomas Lompe, and Martin W. Zwierlein, Quantum-Gas Demonstration of the trapped-ion quantum CCD computer
Microscope for Fermionic Atoms, Phys. Rev. Lett. 114, architecture, Nature 592, 209 (2021).
193001 (2015). [261] K. Wright, et al., Benchmarking an 11-qubit quantum
[248] Florian Schäfer, Takeshi Fukuhara, Seiji Sugawa, Yosuke computer, Nat. Comm. 10, 5464 (2019).
Takasu, and Yoshiro Takahashi, Tools for quantum sim- [262] Justin E. Christensen, David Hucul, Wesley C. Campbell,
ulation with ultracold atoms in optical lattices, Nat. Rev. and Eric R. Hudson, High-fidelity manipulation of a qubit
Phys. 2, 411 (2020). enabled by a manufactured nucleus, npj Quantum Inf. 6,
[249] Giulia Semeghini, Harry Levine, Alexander Keesling, 35 (2020).
Sepehr Ebadi, Tout T. Wang, Dolev Bluvstein, Ruben [263] C. Monroe and J. Kim, Scaling the ion trap quantum
Verresen, Hannes Pichler, Marcin Kalinowski, and Rhine processor, Science 339, 1164 (2013).
Samajdar, et al., Probing topological spin liquids on a pro- [264] C. Monroe, R. Raussendorf, A. Ruthven, K. R. Brown,
grammable quantum simulator, Science 374, 1242 (2021). P. Maunz, L.-M. Duan, and J. Kim, Large-scale modular
[250] Dolev Bluvstein, Ahmed Omran, Harry Levine, Alexan- quantum-computer architecture with atomic memory and
der Keesling, Giulia Semeghini, Sepehr Ebadi, Tout T. photonic interconnects, Phys. Rev. A 89, 022317 (2014).
Wang, Alexios A. Michailidis, Nishad Maskara, and Wen [265] “Values based on recent demonstrations by the DOE’s
Wei Ho, et al., Controlling quantum many-body dynamics Advanced Quantum Testbed at the LBNL,” [Link]
in driven Rydberg atom arrays, Science 371, 1355 (2021). gov/.

027001-58
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

[266] Seth Lloyd, Universal quantum simulators, Science 273, [280] Sarmed A. Rahman, Randy Lewis, Emanuele Mendi-
1073 (1996). celli, and Sarah Powell, SU(2) lattice gauge theory on
[267] Laird Egan, Dripto M. Debroy, Crystal Noel, Andrew a quantum annealer, Phys. Rev. D 104, 034501 (2021).
Risinger, Daiwei Zhu, Debopriyo Biswas, Michael New- ArXiv:2103.08661,
man, Muyuan Li, Kenneth R. Brown, and Marko Cetina, [281] Marc Illa and Martin J. Savage, Basic elements for simula-
et al., Fault-tolerant control of an error-corrected qubit, tions of Standard Model physics with quantum annealers:
Nature 598, 281 (2021). Multigrid and clock states, (2022), ArXiv:2202.12340.
[268] C. Ryan-Anderson, J. G. Bohnet, K. Lee, D. Gresh, A. [282] Ali Hamed Moosavian, James R. Garrison, and Stephen P.
Hankin, J. P. Gaebler, D. Francois, A. Chernoguzov, D. Jordan, Site-by-site quantum state preparation algorithm
Lucchetti, and N. C. Brown, et al., Realization of Real- for preparing vacua of fermionic lattice field theories,
Time Fault-Tolerant Quantum Error Correction, Phys. (2019), arXiv preprint ArXiv:1911.03505.
Rev. X 11, 041058 (2021). [283] Natalie Klco and Martin J. Savage, Minimally entangled
[269] Sebastian Krinner, Nathan Lacroix, Ants Remm, Agustin state preparation of localized wave functions on quantum
Di Paolo, Elie Genois, Catherine Leroux, Christoph computers, Phys. Rev. A 102, 012612 (2020).
Hellings, Stefania Lazar, Francois Swiadek, and Johannes [284] Christian W. Bauer, Plato Deliyannis, Marat Freytsis,
Herrmann, et al., Realizing repeated quantum error cor- and Benjamin Nachman, Practical considerations for the
rection in a distance-three surface code, (2021), arXiv preparation of multivariate Gaussian states on quantum
preprint ArXiv:2112.03708. computers, (2021), arXiv preprint ArXiv:2109.10918.
[270] Reinier W. Heeres, Philip Reinhold, Nissim Ofek, Luigi [285] Zohreh Davoudi, Indrakshi Raychowdhury, and Andrew
Frunzio, Liang Jiang, Michel H. Devoret, and Robert J. Shaw, Search for efficient formulations for Hamiltonian
Schoelkopf, Implementing a universal gate set on a logical simulation of non-Abelian lattice gauge theories, Phys.
qubit encoded in an oscillator, Nat. Commun. 8, 1 (2017). Rev. D 104, 074505 (2021).
[271] Christa Flühmann, Thanh Long Nguyen, Matteo Marinelli, [286] Filip B. Maciejewski, Zoltán Zimborás, and Michał Osz-
Vlad Negnevitsky, Karan Mehta, and J. P. Home, Encod- maniec, Mitigation of readout noise in near-term quantum
ing a qubit in a trapped-ion mechanical oscillator, Nature devices by classical post-processing based on detector
566, 513 (2019). tomography, Quantum 4, 257 (2020).
[272] Nhung H. Nguyen, Muyuan Li, Alaina M. Green, C. [287] Joel J. Wallman and Joseph Emerson, Noise tailoring for
Huerta Alderete, Yingyue Zhu, Daiwei Zhu, Kenneth R. scalable quantum computation via randomized compiling,
Brown, and Norbert M. Linke, Demonstration of Shor Phys. Rev. A 94, 052325 (2016).
Encoding on a Trapped-Ion Quantum Computer, Phys. [288] Ying Li and Simon C. Benjamin, Efficient Variational
Rev. Appl. 16, 024057 (2021). Quantum Simulator Incorporating Active Error Minimiza-
[273] Abhishek Rajput, Alessandro Roggero, and Nathan tion, Phys. Rev. X 7, 021050 (2017).
Wiebe, Quantum error correction with gauge symmetries, [289] Zhenyu Cai and Simon C. Benjamin, Constructing smaller
(2021), arXiv preprint ArXiv:2112.05186. Pauli twirling sets for arbitrary error channels, Sci. Rep. 9,
[274] John Preskill, Quantum computing in the NISQ era and 1 (2019).
beyond, Quantum 2, 79 (2018). [290] Zhenyu Cai, Xiaosi Xu, and Simon C. Benjamin, Mitigat-
[275] Alberto Peruzzo, Jarrod McClean, Peter Shadbolt, Man- ing coherent noise using Pauli conjugation, npj Quantum
Hong Yung, Xiao-Qi Zhou, Peter J Love, Alán Aspuru- Inf. 6, 1 (2020).
Guzik, and Jeremy L. O’brien, A variational eigenvalue [291] Kristan Temme, Sergey Bravyi, and Jay M. Gambetta,
solver on a photonic quantum processor, Nat. Commun. 5, Error Mitigation for Short-Depth Quantum Circuits, Phys.
1 (2014). Rev. Lett. 119, 180509 (2017).
[276] Jarrod R McClean, Jonathan Romero, Ryan Babbush, [292] Abhinav Kandala, Kristan Temme, Antonio D. Córcoles,
and Alán Aspuru-Guzik, The theory of variational hybrid Antonio Mezzacapo, Jerry M. Chow, and Jay M. Gam-
quantum-classical algorithms, New J. Phys. 18, 023023 betta, Error mitigation extends the computational reach of
(2016). a noisy quantum processor, Nature 567, 491 (2019).
[277] Edward Farhi, Jeffrey Goldstone, and Sam Gutmann, [293] Andre He, Benjamin Nachman, Wibe A. de Jong, and
A quantum approximate optimization algorithm, (2014), Christian W. Bauer, Zero-noise extrapolation for quantum-
arXiv preprint ArXiv:1411.4028. gate error mitigation with identity insertions, Phys. Rev. A
[278] Jules Tilly, Hongxiang Chen, Shuxiang Cao, Dario 102, 012426 (2020).
Picozzi, Kanav Setia, Ying Li, Edward Grant, Leonard [294] Jinzhao Sun, Xiao Yuan, Takahiro Tsunoda, Vlatko
Wossnig, Ivan Rungger, and George H. Booth, et al., The Vedral, Simon C. Benjamin, and Suguru Endo, Mitigat-
variational quantum eigensolver: A review of methods and ing Realistic Noise in Practical Noisy Intermediate-Scale
best practices, (2021), arXiv preprint ArXiv:2111.05176. Quantum Devices, Phys. Rev. Appl. 15, 034026 (2021).
[279] Mark Johnson, Mohammad Amin, S. Gildert, Trevor [295] Christian Bauer, Andre He, Wibe A. de Jong, Benjamin
Lanting, F. Hamze, N. Dickson, R. Harris, Andrew Nachman, and Vincent R. Pascuzzi, Computationally effi-
Berkley, Jan Johansson, Paul Bunyk, E. Chapple, C. cient zero noise extrapolation for quantum gate error
Enderud, Jeremy Hilton, Kamran Karimi, E. Ladizin- mitigation, (2021), arXiv preprint ArXiv:2110.13338.
sky, Nicolas Ladizinsky, T. Oh, I. Perminov, C. Rich, [296] Suguru Endo, Simon C. Benjamin, and Ying Li, Practical
and Geordie Rose, Quantum annealing with manufactured Quantum Error Mitigation for Near-Future Applications,
spins, Nature 473, 194 (2011). Phys. Rev. X 8, 031027 (2018).

027001-59
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[297] Suguru Endo, Qi Zhao, Ying Li, and Simon Benjamin, and [318] John B. Kogut, The lattice gauge theory approach to
Xiao Yuan, Mitigating algorithmic errors in a Hamiltonian quantum chromodynamics, Rev. Mod. Phys. 55, 775
simulation, Phys. Rev. A 99, 012334 (2019). (1983).
[298] Matthew Otten and Stephen K. Gray, Recovering noise- [319] Anthony Ciavarella, Algorithm for quantum computation
free quantum observables, Phys. Rev. A 99, 012338 of particle decays, Phys. Rev. D 102, 094505 (2020).
(2019). [320] Michael Kreshchuk, William M. Kirby, Gary Goldstein,
[299] Matthew Otten and Stephen K. Gray, Accounting for Hugo Beauchemin, and Peter J. Love, Quantum simula-
errors in quantum algorithms via individual error reduc- tion of quantum field theory in the light-front formulation,
tion, npj Quantum Inf. 5, 1 (2019). (2020), arXiv preprint ArXiv:2002.04016.
[300] Armands Strikis, Dayue Qin, Yanzhu Chen, Simon C. [321] Adrián Pérez-Salinas, Juan Cruz-Martinez, Abdulla A.
Benjamin, and Ying Li, Learning-Based Quantum Error Alhajri, and Stefano Carrazza, Determining the proton
Mitigation, PRX Quantum 2, 040330 (2021). content with a quantum computer, Phys. Rev. D 103,
[301] Miroslav Urbanek, Benjamin Nachman, Vincent R. Pas- 034027 (2021).
cuzzi, Andre He, Christian W. Bauer, and Wibe A. de [322] Thomas D. Cohen, Henry Lamm, Scott Lawrence, and
Jong, Mitigating Depolarizing Noise on Quantum Com- Yukari Yamauchi, NuQS Collaboration, et al., Quantum
puters with Noise-Estimation Circuits, Phys. Rev. Lett. algorithms for transport coefficients in gauge theories,
127, 270502 (2021). Phys. Rev. D 104, 094514 (2021).
[302] Xavi Bonet-Monroig, Ramiro Sagastizabal, M. Singh, and [323] Roberto Verdel, Fangli Liu, Seth Whitsitt, Alexey V. Gor-
T. E. O’Brien, Low-cost error mitigation by symmetry shkov, and Markus Heyl, Real-time dynamics of string
verification, Phys. Rev. A 98, 062339 (2018). breaking in quantum spin chains, Phys. Rev. B 102,
[303] Matthew Otten, Cristian L. Cortes, and Stephen K. 014308 (2020).
Gray, Noise-resilient quantum dynamics using symmetry- [324] Stefan Höche, in Journeys Through the Precision Fron-
preserving ansatzes, (2019), arXiv preprint ArXiv:1910. tier: Amplitudes for Colliders: TASI 2014 Proceedings
06284. of the 2014 Theoretical Advanced Study Institute in
[304] Ryuji Takagi, Suguru Endo, Shintaro Minagawa, and Mile Elementary Particle Physics (World Scientific, 2016),
Gu, Fundamental limits of quantum error mitigation, npj p. 235.
Quantum Inf. 8, 114 (2022). ArXiv:2109.04457, [325] Bo Andersson, Gösta Gustafson, Gunnar Ingelman, and
[305] Ryuji Takagi, Hiroyasu Tajima, and Mile Gu, Univer- Torbjörn Sjöstrand, Parton fragmentation and string
sal sampling lower bounds for quantum error mitigation, dynamics, Phys. Rep. 97, 31 (1983).
(2022), ArXiv:2208.09178. [326] Josh M. Deutsch, Quantum statistical mechanics in a
[306] Kento Tsubouchi, Takahiro Sagawa, and Nobuyuki Yosh- closed system, Phys. Rev. A 43, 2046 (1991).
ioka, Universal cost bound of quantum error miti- [327] Mark Srednicki, Chaos and quantum thermalization, Phys.
gation based on quantum estimation theory, (2022), Rev. E 50, 888 (1994).
ArXiv:2208.09385. [328] Luigi Amico, Rosario Fazio, Andreas Osterloh, and
[307] Thien Nguyen, Lindsay Bassman, Dmitry Lyakh, Alexan- Vlatko Vedral, Entanglement in many-body systems, Rev.
der McCaskey, Vicente Leyton-Ortega, Raphael Pooser, Mod. Phys. 80, 517 (2008).
Wael Elwasif, Travis S. Humble, and Wibe A. de Jong, [329] Jens Eisert, Marcus Cramer, and Martin B. Plenio, Collo-
in 2021 IEEE 18th International Conference on Software quium: Area laws for the entanglement entropy, Rev. Mod.
Architecture Companion (ICSA-C) (IEEE, 2021), p. 110. Phys. 82, 277 (2010).
[308] “Advanced Quantum Testbed,” [Link] [330] Thomas Pichler, Marcello Dalmonte, Enrique Rico, Peter
[309] “IBM Quantum Computing,” [Link] Zoller, and Simone Montangero, Real-Time Dynamics in
tum-computing/, accessed: 2022-03-24. U(1) Lattice Gauge Theories with Tensor Networks, Phys.
[310] “QuEra Computing Inc,” [Link] acces- Rev. X 6, 011023 (2016).
sed: 2022-03-24. [331] Jad C. Halimeh, Luca Barbiero, Philipp Hauke, Fabian
[311] “Xanadu Quantum Technologies,” [Link] Grusdt, and Annabelle Bohrdt, Robust quantum many-
ai/, accessed: 2022-03-24. body scars in lattice gauge theories, (2022), arXiv preprint
[312] “Quantum Circuits Inc,” [Link] ArXiv:2203.08828.
accessed: 2022-03-24. [332] Debasish Banerjee and Arnab Sen, Quantum Scars From
[313] “D-Wave Systems Inc,” [Link] Zero Modes in an Abelian Lattice Gauge Theory on
accessed: 2022-03-24. Ladders, Phys. Rev. Lett. 126, 220601 (2021).
[314] “USQCD Collaboration,” [Link] [333] Marlon Brenes, Marcello Dalmonte, Markus Heyl, and
[Link]. Antonello Scardicchio, Many-Body Localization Dynam-
[315] “National QIS Research Centers,” [Link] ics from Gauge Invariance, Phys. Rev. Lett. 120, 030601
Initiatives/QIS/QIS-Centers. (2018).
[316] Ciaran Hughes, Doug Finke, Dan-Adrian German, Celia [334] Marcello Dalmonte, Viktor Eisler, Marco Falconi, and
Merzbacher, Patrick M. Vora, and H. J. Lewandowski, Benoıt Vermersch, Entanglement hamiltonians: From
Assessing the needs of the quantum industry, (2021), field theory, to lattice models and experiments, (2022),
arXiv preprint ArXiv:2109.03601. arXiv preprint ArXiv:2202.05045.
[317] John B. Kogut, An introduction to lattice gauge theory and [335] Christian Kokail, Rick van Bijnen, Andreas Elben, Benoît
spin systems, Rev. Mod. Phys. 51, 659 (1979). Vermersch, and Peter Zoller, Entanglement hamiltonian

027001-60
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

tomography in quantum simulation, Nat. Phys. 17, 936 [351] Alessandro Roggero, Entanglement and many-body
(2021). effects in collective neutrino oscillations, Phys. Rev. D
[336] Hannes Pichler, Guanyu Zhu, Alireza Seif, Peter Zoller, 104, 103016 (2021).
and Mohammad Hafezi, Measurement Protocol for the [352] Alessandro Roggero, Dynamical phase transitions in mod-
Entanglement Spectrum of Cold Atoms, Phys. Rev. X 6, els of collective neutrino oscillations, Phys. Rev. D 104,
041033 (2016). 123023 (2021).
[337] Justin Yirka and Yiğit Subaşı, Qubit-efficient entangle- [353] Michael J. Cervia, Pooja Siwach, Amol V. Patwardhan,
ment spectroscopy using qubit resets, Quantum 5, 535 A. B. Balantekin, S. N. Coppersmith, and Calvin W. John-
(2021). son, Collective neutrino oscillations with tensor networks
[338] Christian Kokail, Bhuvanesh Sundar, Torsten V. Zache, using a time-dependent variational principle, Phys. Rev. D
Andreas Elben, Benoít Vermersch, Marcello Dalmonte, 105, 123025 (2022).
Rick van Bijnen, and Peter Zoller, Quantum Variational [354] Noemi Rocco, Ab initio calculations of lepton-nucleus
Learning of the Entanglement Hamiltonian, Phys. Rev. scattering, Front. Phys. 8, 116 (2020).
Lett. 127, 170501 (2021). [355] Joshua Isaacson, Stefan Höche, Diego Lopez Gutierrez,
[339] Aditi Mitra, Quantum quench dynamics, Annu. Rev. Con- and Noemi Rocco, A novel event generator for the auto-
dens. Matter Phys. 9, 245 (2018). mated simulation of neutrino scattering, (2021), arXiv
[340] Elena Canovi, Davide Rossini, Rosario Fazio, Giuseppe E. preprint ArXiv:2110.15319.
Santoro, and Alessandro Silva, Quantum quenches, ther- [356] Noemi Rocco, Winfried Leidemann, Alessandro Lovato,
malization, and many-body localization, Phys. Rev. B 83, and Giuseppina Orlandini, Relativistic effects in ab initio
094431 (2011). electron-nucleus scattering, Phys. Rev. C. Nucl. Phys. 97,
[341] Hannes Bernien, Sylvain Schwartz, Alexander Keesling, 055501 (2018).
Harry Levine, Ahmed Omran, Hannes Pichler, Soonwon [357] Steven Weinberg, Cosmological Production of Baryons,
Choi, Alexander S. Zibrov, Manuel Endres, and Markus Phys. Rev. Lett. 42, 850 (1979).
Greiner, et al., Probing many-body dynamics on a 51-atom [358] Ian Affleck and Michael Dine, A new mechanism for
quantum simulator, Nature 551, 579 (2017). baryogenesis, Nucl. Phys. B 249, 361 (1985).
[342] Jiehang Zhang, Guido Pagano, Paul W. Hess, Anto- [359] Mariano Quiros, Finite temperature field theory and phase
nis Kyprianidis, Patrick Becker, Harvey Kaplan, Alexey transitions, Proc., Summer School High-Energy Phys.
V. Gorshkov, Z.-X. Gong, and Christopher Monroe, Cosmol.: Trieste, Italy 1999, 187 (1998).
Observation of a many-body dynamical phase transi- [360] Mark Hindmarsh, Marvin Lüben, Johannes Lumma, and
tion with a 53-qubit quantum simulator, Nature 551, 601 Martin Pauly, Phase transitions in the early universe,
(2017). SciPost Phys. Lecture Notes , ().
[343] Scott Aaronson, Shadow tomography of quantum states, [361] David Curtin, Patrick Meade, and Harikrishnan Ramani,
SIAM J. Comput. 49, STOC18–368 (2019). Thermal resummation and phase transitions, Eur. Phys. J.
[344] Hsin-Yuan Huang, Richard Kueng, and John Preskill, Pre- C 78, 1 (2018).
dicting many properties of a quantum system from very [362] King Lun Ng, Bogdan Opanchuk, Manushan Thenabadu,
few measurements, Nat. Phys. 16, 1050 (2020). Margaret Reid, and Peter D. Drummond, Fate of the False
[345] Fernando G. S. L. Brandao and Aram W. Harrow, Product- Vacuum: Finite Temperature, Entropy, and Topological
state approximations to quantum states, Commun. Math. Phase in Quantum Simulations of the Early Universe,
Phys. 342, 47 (2016). PRX Quantum 2, 010350 (2021).
[346] Benjamin Hall, Alessandro Roggero, Alessandro Baroni, [363] Bipasha Chakraborty, Masazumi Honda, Taku Izubuchi,
and Joseph Carlson, Simulation of collective neutrino Yuta Kikuchi, and Akio Tomiya, Digital quantum sim-
oscillations on a quantum computer, Phys. Rev. D 104, ulation of the Schwinger model with topological term
063009 (2021). via adiabatic state preparation, (2020), arXiv preprint
[347] Kübra Yeter-Aydeniz, Shikha Bangar, George Siopsis, ArXiv:2001.00485.
and Raphael C. Pooser, Collective neutrino oscilla- [364] Angus Kan, Lena Funcke, Stefan Kühn, Luca Dellantonio,
tions on a quantum computer, (2021), arXiv preprint Jinglei Zhang, Jan F. Haase, Christine A. Muschik, and
ArXiv:2104.03273. Karl Jansen, Investigating a (3+1) D topological θ-term
[348] Michael J. Cervia, Amol V. Patwardhan, A. B. Balantekin, in the Hamiltonian formulation of lattice gauge theories
S. N. Coppersmith, and Calvin W. Johnson, Entanglement for quantum and classical simulations, Phys. Rev. D 104,
and collective flavor oscillations in a dense neutrino gas, 034504 (2021).
Phys. Rev. D 100, 083001 (2019). [365] Xizhi Han and Sean A. Hartnoll, et al., Deep Quantum
[349] Amol V. Patwardhan, Michael J. Cervia, and A. B. Bal- Geometry of Matrices, Phys. Rev. X 10, 011069 (2020).
antekin, Spectral splits and entanglement entropy in col- [366] Enrico Rinaldi, Xizhi Han, Mohammad Hassan, Yuan
lective neutrino oscillations, Phys. Rev. D 104, 123035 Feng, Franco Nori, Michael McGuigan, and Masanori
(2021). Hanada, Matrix-Model Simulations using Quantum Com-
[350] Joshua D. Martin, A. Roggero, Huaiyu Duan, J. Carlson, puting, Deep Learning, and Lattice Monte Carlo, PRX
and V. Cirigliano, Classical and quantum evolution in a Quantum 3, 010324 (2022).
simple coherent neutrino problem, (2021), arXiv preprint [367] Katrin Becker and Melanie Becker, A two-loop test of
ArXiv:2112.12686. m(atrix) theory, Nucl. Phys. B 506, 48 (1997).

027001-61
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[368] Katrin Becker, Melanie Becker, Joseph Polchinski, and [386] S. R. Beane, W. Detmold, Konstantinos Orginos, and M.
Arkady Tseytlin, Higher order graviton scattering in J. Savage, Nuclear physics from lattice QCD, Prog. Part.
m(atrix) theory, Phys. Rev. D 56, R3174 (1997). Nucl. Phys. 66, 1 (2011).
[369] Yuji Okawa and Tamiaki Yoneya, Multi-body interactions [387] Zohreh Davoudi, in International Conference on Few-
of d-particles in supergravity and matrix theory, Nucl. Body Problems in Physics, (Springer, 2018), p. 597.
Phys. B 538, 67 (1999). [388] Richard C. Brower, Anna Hasenfratz, Ethan T. Neil,
[370] Hrant Gharibyan, Masanori Hanada, Masazumi Honda, Simon Catterall, George Fleming, Joel Giedt, Enrico
and Junyu Liu, Toward simulating superstring/m-theory Rinaldi, David Schaich, Evan Weinberg, and Oliver
on a quantum computer, J. High Energy Phys. 2021, 1 Witzel, Lattice gauge theory for physics beyond the stan-
(2021). dard model, Eur. Phys. J. A 55, 1 (2019).
[371] Ping Gao, Daniel Louis Jafferis, and Aron C. Wall, [389] Kenneth G. Wilson, Confinement of quarks, Phys. Rev. D
Traversable wormholes via a double trace deformation, J. 10, 2445 (1974).
High Energy Phys. 2017, 1 (2017). [390] Natalie Klco and Martin J. Savage, Digitization of scalar
[372] Juan Maldacena, Douglas Stanford, and Zhenbin Yang, fields for quantum computing, Phys. Rev. A 99, 052335
Diving into traversable wormholes, Fortschritte der Phys. (2019).
65, 1700034 (2017). [391] Yao Ji, Henry Lamm, and Shuchen Zhu, NuQS Collab-
[373] Adam R. Brown, Hrant Gharibyan, Stefan Leichenauer, oration, et al., Gluon field digitization via group space
Henry W. Lin, Sepehr Nezami, Grant Salton, Leonard decimation for quantum computers, Phys. Rev. D 102,
Susskind, Brian Swingle, and Michael Walter, Quantum 114513 (2020).
gravity in the lab: Teleportation by size and traversable [392] Yao Ji, and Henry Lamm, and Shuchen Zhu, Gluon dig-
wormholes, (2019), arXiv preprint ArXiv:1911.06314. itization via character expansion for quantum computers,
[374] Sepehr Nezami, Henry W. Lin, Adam R. Brown, Hrant (2022), arXiv preprint ArXiv:2203.02330.
Gharibyan, Stefan Leichenauer, Grant Salton, Leonard [393] Marcela Carena, Henry Lamm, Ying-Ying Li, and Wan-
Susskind, Brian Swingle, and Michael Walter, Quan- qiang Liu, Lattice renormalization of quantum simula-
tum gravity in the lab: Teleportation by size and tions, Phys. Rev. D 104, 094519 (2021).
traversable wormholes, part ii, (2021), arXiv preprint [394] Marcela Carena, Henry Lamm, Ying-Ying Li, and Wan-
ArXiv:2102.01064. qiang Liu, Improved hamiltonians for quantum simula-
[375] Thomas Schuster, Bryce Kobrin, Ping Gao, Iris Cong, tions, (2022), arXiv preprint ArXiv:2203.02823.
Emil T. Khabiboulline, Norbert M. Linke, Mikhail D. [395] Daniel C. Hackett, Kiel Howe, Ciaran Hughes, William
Lukin, Christopher Monroe, Beni Yoshida, and Norman Jay, Ethan T. Neil, and James N. Simone, Digitizing gauge
Y. Yao, Many-body quantum teleportation via operator fields: Lattice Monte Carlo results for future quantum
spreading in the traversable wormhole protocol, (2021), computers, Phys. Rev. A 99, 062341 (2019).
arXiv preprint ArXiv:2102.00010. [396] Andrei Alexandru, Paulo F. Bedaque, Siddhartha Har-
[376] Csaba Csáki and Philip Tanedo, Beyond the standard malkar, Henry Lamm, Scott Lawrence, and Neill C. War-
model, (2016), arXiv preprint ArXiv:1602.04228. rington, NuQS Collaboration, et al., Gluon field digitiza-
[377] Scott Willenbrock and Cen Zhang, Effective field theory tion for quantum computers, Phys. Rev. D 100, 114501
beyond the standard model, Annu. Rev. Nucl. Part. Sci. (2019).
64, 83 (2014). [397] Andrei Alexandru, Paulo F. Bedaque, Ruairí Brett, and
[378] Kalliopi Petraki and Raymond R. Volkas, Review of Henry Lamm, The spectrum of qubitized QCD: Glue-
asymmetric dark matter, Int. J. Mod. Phys. A 28, 1330028 balls in a S(1080) gauge theory, (2021), arXiv preprint
(2013). ArXiv:2112.08482.
[379] Graham D. Kribs and Ethan T. Neil, Review of strongly- [398] Tobias Hartung, Timo Jakobs, Karl Jansen, Johann Ost-
coupled composite dark matter models and lattice simula- meyer, and Carsten Urbach, Digitising SU(2) gauge
tions, Int. J. Mod. Phys. A 31, 1643004 (2016). fields and the freezing transition, (2022), arXiv preprint
[380] Andreas Albert, Martin Bauer, Jim Brooke, Oliver Buch- ArXiv:2201.09625.
mueller, David G. Cerdeno, Matthew Citron, Gavin [399] Danny Paulson, Luca Dellantonio, Jan F. Haase, Alessio
Davies, Annapaola de Cosa, Albert De Roeck, and Andrea Celi, Angus Kan, Andrew Jena, Christian Kokail, Rick
De Simone, et al., Towards the next generation of sim- van Bijnen, Karl Jansen, and Peter Zoller, et al., Simu-
plified dark matter models, Phys. Dark Universe 16, 49 lating 2D Effects in Lattice Gauge Theories on a Quantum
(2017). Computer, PRX Quantum 2, 030334 (2021).
[381] L. Arturo Ureña-López, Brief review on scalar field dark [400] Yannick Meurice, Examples of symmetry-preserving trun-
matter models, Front. Astron. Space Sci., 47 (2019). cations in tensor field theory, Phys. Rev. D 100, 014506
[382] Kenneth G. Wilson, The origins of lattice gauge theory, (2019).
(2004), arXiv preprint ArXiv:hep-lat/0412043. [401] Yannick Meurice, Discrete aspects of continuous sym-
[383] István Montvay and Gernot Münster, Quantum Fields on metries in the tensorial formulation of Abelian gauge
a Lattice (Cambridge University Press, 1997). theories, Phys. Rev. D 102, 014506 (2020).
[384] Heinz J. Rothe, Lattice Gauge Theories: An Introduction [402] Y. Meurice, Accurate exponents from approximate tensor
(World Scientific Publishing Company, 2012). renormalizations, Phys. Rev. B 87, 064422 (2013).
[385] Claudio Rebbi, Lattice Gauge Theories and Monte Carlo [403] Alan Denbleyker, Yuzhi Liu, Y. Meurice, M. P. Qin, T.
Simulations (World Scientific, 1983). Xiang, Z. Y. Xie, J. F. Yu, and Haiyuan Zou, Controlling

027001-62
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

sign problems in spin models using tensor renormaliza- [419] Muhammad Asaduzzaman, Simon Catterall, and Judah
tion, Phys. Rev. D 89, 016008 (2014). Unmuth-Yockey, Tensor network formulation of two-
[404] Hikaru Kawauchi and Shinji Takeda, Phase structure anal- dimensional gravity, Phys. Rev. D 102, 054510 (2020).
ysis of CP (n − 1) model using tensor renormalization [420] Jacques Bloch, Raghav G. Jha, Robert Lohmayer,
group, (2016), arXiv preprint ArXiv:1611.00921. and Maximilian Meister, Tensor renormalization group
[405] Katsumasa Nakayama, Lena Funcke, Karl Jansen, Ying- study of the 3d o(2) model, (2021), arXiv preprint
Jer Kao, and Stefan Kühn, Phase structure of the CP (1) ArXiv:2105.08066.
model in the presence of a topological θ-term, Phys. Rev. [421] Xiao Yuan, Jinzhao Sun, Junyu Liu, Qi Zhao, and You
D 105, 054507 (2022). Zhou, Quantum Simulation with Hybrid Tensor Networks,
[406] Haiyuan Zou, Yuzhi Liu, Chen-Yen Lai, J. Unmuth- Phys. Rev. Lett. 127, 040501 (2021).
Yockey, Li-Ping Yang, A. Bazavov, Z. Y. Xie, T. Xiang, S. [422] These activities are detailed in a Snowmass whitepaper
Chandrasekharan, and S.-W. Tsai, et al., Progress towards [634] focused on the tensor approach to QFTs.
quantum simulating the classical O(2) model, Phys. Rev. [423] Philip D. Mannheim, Peter Lowdon, and Stanley J. Brod-
A 90, 063603 (2014). sky, Comparing light-front quantization with instant-time
[407] Alexei Bazavov, Yannick Meurice, S.-W. Tsai, Judah quantization, Phys. Rep. 891, 1 (2021).
Unmuth-Yockey, and Jin Zhang, Gauge-invariant imple- [424] Junyu Liu and Yuan Xin, Quantum simulation of quantum
mentation of the Abelian-Higgs model on optical lattices, field theories as quantum chemistry, J. High Energy Phys.
Phys. Rev. D 92, 076003 (2015). 2020, 1 (2020).
[408] Jin Zhang, Judah Unmuth-Yockey, Johannes Zeiher, [425] Kenneth G. Wilson, Timothy S. Walhout, Avaroth
Alexei Bazavov, S.-W. Tsai, and Yannick Meurice, Quan- Harindranath, Wei-Min Zhang, Robert J. Perry, and
tum Simulation of the Universal Features of the Polyakov Stanisław D. Głazek, Nonperturbative QCD: A weak-
Loop, Phys. Rev. Lett. 121, 223201 (2018). coupling treatment on the light front, Phys. Rev. D 49,
[409] Yannick Meurice, Theoretical methods to design and test 6720 (1994).
quantum simulators for the compact abelian Higgs model, [426] Sofia S. Chabysheva and John R. Hiller, Zero momentum
Phys. Rev. D 104, 094513 (2021). modes in discrete light-cone quantization, Phys. Rev. D
[410] Yuya Shimizu and Yoshinobu Kuramashi, Grassmann ten- 79, 096012 (2009).
sor renormalization group approach to one-flavor lattice [427] L’ubomír Martinovič and Alexander Dorokhov, Vacuum
Schwinger model, Phys. Rev. D 90, 014508 (2014). loops in light-front field theory, Physics Letters B 811,
[411] Yuya Shimizu and Yoshinobu Kuramashi, Critical behav- 135925 (2020).
ior of the lattice Schwinger model with a topological [428] Sophia S. Chabysheva and John R. Hiller, Tadpoles and
term at θ = π using the Grassmann tensor renormalization vacuum bubbles in light-front quantization, (2022), arXiv
group, Phys. Rev. D 90, 074503 (2014). preprint ArXiv:2201.00123.
[412] Shinji Takeda and Yusuke Yoshimura, Grassmann ten- [429] Weijie Du, James P. Vary, Xingbo Zhao, and Wei Zuo,
sor renormalization group for the one-flavor lattice Quantum simulation of nuclear inelastic scattering, Phys.
Gross–Neveu model with finite chemical potential, Prog. Rev. A 104, 012611 (2021).
Theor. Exp. Phys. 2015, 043B01 (2015). [430] Stephan Caspar and Hersh Singh, From asymptotic free-
[413] Ryo Sakai, Shinji Takeda, and Yusuke Yoshimura, dom to θ vacua: Qubit embeddings of the O(3) nonlinear
Higher-order tensor renormalization group for relativistic σ model, (2022), arXiv preprint ArXiv:2203.15766.
fermion systems, Prog. Theor. Exp. Phys. 2017, 063B07 [431] Hanqing Liu and Shailesh Chandrasekharan, Qubit reg-
(2017). ularization and qubit embedding algebras, Symmetry 14,
[414] Daisuke Kadoh, Yoshinobu Kuramashi, Yoshifumi Naka- 305 (2022).
mura, Ryo Sakai, Shinji Takeda, and Yusuke Yoshimura, [432] Jiayu Shen, Di Luo, Chenxi Huang, Bryan K. Clark, Aida
Tensor network formulation for two-dimensional lattice X. El-Khadra, Bryce Gadway, and Patrick Draper, Sim-
N = 1 Wess-Zumino model, J. High Energy Phys. 2018, ulating quantum mechanics with a θ-term and ‘t Hooft
1 (2018). anomaly on a synthetic dimension, (2021), arXiv preprint
[415] Daisuke Kadoh, Yoshinobu Kuramashi, Yoshifumi Naka- ArXiv:2107.08073.
mura, Ryo Sakai, Shinji Takeda, and Yusuke Yoshimura, [433] Alexander J. Buser, Hrant Gharibyan, Masanori Hanada,
Tensor network analysis of critical coupling in two dimen- Masazumi Honda, and Junyu Liu, Quantum simulation of
sional φ4 theory, J. High Energy Phys. 2019, 1 (2019). gauge theory via orbifold lattice, J. High Energy Phys.
[416] Daiki Adachi, Tsuyoshi Okubo, and Synge Todo, 2021, 1 (2021).
Anisotropic tensor renormalization group, Phys. Rev. B [434] Judah F. Unmuth-Yockey, Gauge-invariant rotor Hamilto-
102, 054432 (2020). nian from dual variables of 3D U(1) gauge theory, Phys.
[417] Shinichiro Akiyama, Daisuke Kadoh, Yoshinobu Kura- Rev. D 99, 074502 (2019).
mashi, Takumi Yamashita, and Yusuke Yoshimura, Tensor [435] Julian Bender and Erez Zohar, Gauge redundancy-free
renormalization group approach to four-dimensional com- formulation of compact QED with dynamical matter for
plex φ4 theory at finite density, J. High Energy Phys. quantum and classical computations, Phys. Rev. D 102,
2020, 1 (2020). 114517 (2020).
[418] Yoshinobu Kuramashi and Yusuke Yoshimura, Three- [436] C. J. Hamer, Zheng Weihong, and J. Oitmaa, Series expan-
dimensional finite temperature Z2 gauge theory with ten- sions for the massive Schwinger model in hamiltonian
sor network scheme, J. High Energy Phys. 2019, 1 (2019). lattice theory, Phys. Rev. D 56, 55 (1997).

027001-63
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[437] Pablo Sala, Tao Shi, Stefan Kühn, Mari Carmen Banuls, [455] Yu Tong, Victor V. Albert, Jarrod R. McClean, John
Eugene Demler, and Juan Ignacio Cirac, Variational study Preskill, and Yuan Su, Provably accurate simulation
of U(1) and SU(2) lattice gauge theories with Gaussian of gauge theories and bosonic systems, (2021), arXiv
states in 1+1 dimensions, Phys. Rev. D 98, 034505 (2018). preprint ArXiv:2110.06942.
[438] Raúl A. Briceño, Juan V. Guerrero, Maxwell T. Hansen, [456] Rolando D. Somma, Quantum simulations of one
and Alexandru M. Sturzu, Role of boundary conditions dimensional quantum systems, (2015), arXiv preprint
in quantum computations of scattering observables, Phys. ArXiv:1503.06319.
Rev. D 103, 014506 (2021). [457] Alexandru Macridin, Panagiotis Spentzouris, James
[439] Masuo Suzuki, General theory of fractal path integrals Amundson, and Roni Harnik, Digital quantum computa-
with applications to many-body theories and statistical tion of fermion-boson interacting systems, Phys. Rev. A
physics, J. Math. Phys. 32, 400 (1991). 98, 042312 (2018).
[440] Nathan Wiebe, Dominic Berry, Peter Hoyer, and Barry C. [458] Alexandru Macridin, Panagiotis Spentzouris, James
Sanders, Higher order decompositions of ordered operator Amundson, and Roni Harnik, Electron-Phonon Systems
exponentials, J. Phys. A: Math. Theor. 43, 065203 (2010). on a Universal Quantum Computer, Phys. Rev. Lett. 121,
[441] Andrew M. Childs, Yuan Su, Minh C. Tran, Nathan 110504 (2018).
Wiebe, and Shuchen Zhu, Theory of Trotter Error with [459] Nicolas P. D. Sawaya, Tim Menke, Thi Ha Kyaw, Sonika
Commutator Scaling, Phys. Rev. X 11, 011020 (2021). Johri, Alán Aspuru-Guzik, and Gian Giacomo Guerreschi,
[442] Andrew M. Childs and Nathan Wiebe, Hamiltonian sim- Resource-efficient digital quantum simulation of D-level
ulation using linear combinations of unitary operations, systems for photonic, vibrational, and spin-S hamiltoni-
Quantum Inf. Comput. 12, 901 (2012). ans, npj Quantum Inf. 6, 1 (2020).
[443] Dominic W. Berry, Andrew M. Childs, Richard Cleve, [460] P. Jordan and E. Wigner, Über das paulische äquivalen-
Robin Kothari, and Rolando D. Somma, Simulating zverbot, Z. für Phys. 47, 631 (1928).
Hamiltonian Dynamics with a Truncated Taylor Series, [461] Sergey B. Bravyi and Alexei Yu. Kitaev, Fermionic quan-
Phys. Rev. Lett. 114, 090502 (2015). tum computation, Ann. Phys. 298, 210 (2002).
[444] Guang Hao Low and Isaac L. Chuang, Optimal Hamil- [462] F. Verstraete and J. I. Cirac, Mapping local hamiltonians
tonian Simulation by Quantum Signal Processing, Phys. of fermions to local hamiltonians of spins, J. Stat. Mech.
Rev. Lett. 118, 010501 (2017). 2005, P09012 (2005).
[445] Guang Hao Low and Isaac L. Chuang, Hamiltonian simu- [463] R. C. Ball, Fermions Without Fermion Fields, Phys. Rev.
lation by qubitization, Quantum 3, 163 (2019). Lett. 95, 176407 (2005).
[446] Shantanav Chakraborty, András Gilyén, and Stacey [464] James D. Whitfield, Vojtěch Havlíček, and Matthias
Jeffery, The power of block-encoded matrix powers: Troyer, Local spin operators for fermion simulations,
improved regression techniques via faster hamiltonian Phys. Rev. A 94, 030301 (2016).
simulation, (2018), arXiv preprint ArXiv:1804.01973. [465] Mark Steudtner and Stephanie Wehner, Quantum codes
[447] András Gilyén, Yuan Su Guang Hao Low, and Nathan for quantum simulation of fermions on a square lattice of
Wiebe, in Proceedings of the 51st Annual ACM SIGACT qubits, Phys. Rev. A 99, 022308 (2019).
Symposium on Theory of Computing (2019), p. 193. [466] Kanav Setia, Sergey Bravyi, Antonio Mezzacapo, and
[448] Amir Kalev and Itay Hen, Quantum algorithm for sim- James D. Whitfield, Superfast encodings for fermionic
ulating hamiltonian dynamics with an off-diagonal series quantum simulation, Phys. Rev. Res. 1, 033033 (2019).
expansion, Quantum 5, 426 (2021). [467] Yu-An Chen, Exact bosonization in arbitrary dimensions,
[449] Stephen P. Jordan, Keith S. M. Lee, and John Preskill, Phys. Rev. Res. 2, 033527 (2020).
Quantum algorithms for fermionic quantum field theories, [468] Riley W. Chien and James D. Whitfield, Custom fermionic
(2014), arXiv preprint ArXiv:1404.7115. codes for quantum simulation, (2020), arXiv preprint
[450] Erik Gustafson, Yingyue Zhu, Patrick Dreher, Norbert M. ArXiv:2009.11860.
Linke, and Yannick Meurice, Real-time quantum calcula- [469] Charles Derby, Joel Klassen, Johannes Bausch, and Toby
tions of phase shifts using wave packet time delays, Phys. Cubitt, Compact fermion to qubit mappings, Phys. Rev. B
Rev. D 104, 054507 (2021). 104, 035118 (2021).
[451] Mohsen Bagherimehrab, Yuval R. Sanders, Dominic W. [470] Simon V. Mathis, Guglielmo Mazzola, and Ivano Taver-
Berry, Gavin K. Brennen, and Barry C. Sanders, Nearly nelli, Toward scalable simulations of lattice gauge theories
optimal quantum algorithm for generating the ground state on quantum computers, Phys. Rev. D 102, 094501 (2020).
of a free quantum field theory, (2021), arXiv preprint [471] Dave Wecker, Matthew B. Hastings, Nathan Wiebe, Bryan
ArXiv:2110.05708. K. Clark, Chetan Nayak, and Matthias Troyer, Solving
[452] Dean Lee, Joey Bonitati, Gabriel Given, Caleb Hicks, strongly correlated electron models on a quantum com-
Ning Li, Bing-Nan Lu, Abudit Rai, Avik Sarkar, and puter, Phys. Rev. A 92, 062318 (2015).
Jacob Watkins, Projected cooling algorithm for quantum [472] Morten Kjaergaard, Mollie E. Schwartz, Jochen
computation, Phys. Lett. B 807, 135536 (2020). Braumüller, Philip Krantz, Joel I.-J. Wang, Simon Gus-
[453] Erik J. Gustafson, Projective cooling for the transverse tavsson, and William D. Oliver, Superconducting qubits:
Ising model, Phys. Rev. D 101, 071504 (2020). Current state of play, Annu. Rev. Condens. Matter Phys.
[454] Ali Hamed Moosavian and Stephen Jordan, Faster quan- 11, 369 (2020).
tum algorithm to simulate fermionic quantum field theory, [473] Philip Krantz, Morten Kjaergaard, Fei Yan, Terry P.
Phys. Rev. A 98, 012332 (2018). Orlando, Simon Gustavsson, and William D. Oliver,

027001-64
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

A quantum engineer’s guide to superconducting qubits, [490] Ulrik L. Andersen, Jonas S. Neergaard-Nielsen, Peter
Appl. Phys. Rev. 6, 021318 (2019). Van Loock, and Akira Furusawa, Hybrid discrete-and
[474] Ananda Roy and Hubert Saleur, Quantum electronic cir- continuous-variable quantum information, Nat. Phys. 11,
cuit simulation of generalized Sine-Gordon models, Phys. 713 (2015).
Rev. B 100, 155425 (2019). [491] Han-Sen Zhong, Hui Wang, Yu-Hao Deng, Ming-Cheng
[475] Ananda Roy, Dirk Schuricht, Johannes Hauschild, Frank Chen, Li-Chao Peng, Yi-Han Luo, Jian Qin, Dian Wu,
Pollmann, and Hubert Saleur, The quantum Sine-Gordon Xing Ding, and Yi Hu, et al., Quantum computational
model with quantum circuits, Nucl. Phys. B 968, 115445 advantage using photons, Science 370, 1460 (2020).
(2021). [492] J. M. Arrazola, V. Bergholm, K. Brádler, T. R. Bromley,
[476] Jinglei Zhang, Ryan Ferguson, Stefan Kühn, Jan F. M. J. Collins, I. Dhand, A. Fumagalli, T. Gerrits, A. Gous-
Haase, C. M. Wilson, Karl Jansen, and Christine A. sev, and L. G. Helt, et al., Quantum circuits with many
Muschik, Simulating gauge theories with variational photons on a programmable nanophotonic chip, Nature
quantum eigensolvers in superconducting microwave cav- 591, 54 (2021).
ities, (2021), arXiv preprint ArXiv:2108.08248. [493] Mario Motta, Chong Sun, Adrian T. K. Tan, Matthew J.
[477] Di Luo, Jiayu Shen, Michael Highman, Bryan K. Clark, O’Rourke, Erika Ye, Austin J. Minnich, Fernando G. S. L.
Brian DeMarco, Aida X. El-Khadra, and Bryce Gad- Brandão, and Garnet Kin-Lic Chan, Determining eigen-
way, Framework for simulating gauge theories with states and thermal states on a quantum computer using
dipolar spin systems, Phys. Rev. A 102, 032617 quantum imaginary time evolution, Nat. Phys. 16, 205
(2020). (2020).
[478] U.-J. Wiese, Ultracold quantum gases and lattice systems: [494] Kübra Yeter-Aydeniz, Eleftherios Moschandreou, and
Quantum simulation of lattice gauge theories, Ann. Phys. George Siopsis, Quantum imaginary-time evolution
525, 777 (2013). algorithm for quantum field theories with continuous vari-
[479] Enrique Rico, Thomas Pichler, Marcello Dalmonte, Peter ables, Phys. Rev. A 105, 012412 (2022).
Zoller, and Simone Montangero, Tensor Networks for Lat- [495] Shi-Ning Sun, Mario Motta, Ruslan N. Tazhigulov, Adrian
tice Gauge Theories and Atomic Quantum Simulation, T. K. Tan, Garnet Kin-Lic Chan, and Austin J. Min-
Phys. Rev. Lett. 112, 201601 (2014). nich, Quantum Computation of Finite-Temperature Static
[480] Valentin Kasper, Florian Hebenstreit, Markus K. and Dynamical Properties of Spin Systems using Quan-
Oberthaler, and Jürgen Berges, Schwinger pair production tum Imaginary Time Evolution, PRX Quantum 2, 010317
with ultracold atoms, Phys. Lett. B 760, 742 (2016). (2021).
[481] Alexander Mil, Torsten V. Zache, Apoorva Hegde, Andy [496] Jing Wu and Quntao Zhuang, Continuous-Variable Error
Xia, Rohit P. Bhatt, Markus K. Oberthaler, Philipp Hauke, Correction for General Gaussian Noises, Phys. Rev. Appl.
Jürgen Berges, and Fred Jendrzejewski, A scalable real- 15, 034073 (2021).
ization of local U(1) gauge invariance in cold atomic [497] Kosuke Fukui, Akihisa Tomita, and Atsushi Okamoto,
mixtures, Science 367, 1128 (2020). Analog Quantum Error Correction with Encoding a Qubit
[482] Erez Zohar, Alessandro Farace, Benni Reznik, and J. Igna- into an Oscillator, Phys. Rev. Lett. 119, 180507 (2017).
cio Cirac, Digital lattice gauge theories, Phys. Rev. A 95, [498] Ilan Tzitrin, J. Eli Bourassa, Nicolas C. Menicucci, and
023604 (2017). Krishna Kumar Sabapathy, Progress towards practical
[483] Julian Bender, Erez Zohar, Alessandro Farace, and J. Igna- qubit computation using approximate Gottesman-Kitaev-
cio Cirac, Digital quantum simulation of lattice gauge Preskill codes, Phys. Rev. A 101, 032315 (2020).
theories in three spatial dimensions, New J. Phys. 20, [499] L. Tagliacozzo, A. Celi, P. Orland, M. W. Mitchell, and
093001 (2018). M. Lewenstein, Simulation of non-Abelian gauge theories
[484] Yuchen Wang, Zixuan Hu, Barry C. Sanders, and Sabre with optical lattices, Nat. Commun. 4, 1 (2013).
Kais, Qudits and high-dimensional quantum computing, [500] Aram W. Harrow, Avinatan Hassidim, and Seth Lloyd,
Front. Phys. 8, 479 (2020). Quantum Algorithm for Linear Systems of Equations,
[485] Erik J. Gustafson, Prospects for simulating a qudit-based Phys. Rev. Lett. 103, 150502 (2009).
model of 1+1 D scalar QED, Phys. Rev. D 103, 114505 [501] David Simmons-Duffin, A semidefinite program solver for
(2021). the conformal bootstrap, J. High Energy Phys. 2015, 1
[486] Doga Murat Kurkcuoglu, M. Sohaib Alam, Andy C. Y. (2015).
Li, Alexandru Macridin, and Gabriel N. Perdue, Quantum [502] Ning Bao and Junyu Liu, Quantum algorithms for confor-
simulation of φ 4 theories in qudit systems, (2021), arXiv mal bootstrap, Nucl. Phys. B 946, 114702 (2019).
preprint ArXiv:2108.13357. [503] Fernando G. S. L. Brandão, Amir Kalev, Tongyang Li,
[487] Seth Lloyd and Samuel L. Braunstein, in Quantum infor- Cedric Yen-Yu Lin, Krysta M. Svore, and Xiaodi Wu,
mation with continuous variables (Springer, 1999), p. Quantum SDP solvers: Large speed-ups, optimality, and
9. applications to quantum learning, (2017), arXiv preprint
[488] Daniel Gottesman, Alexei Kitaev, and John Preskill, ArXiv:1710.02581.
Encoding a qubit in an oscillator, Phys. Rev. A 64, 012310 [504] Erik Gustafson, Burt Holzman, James Kowalkowski,
(2001). Henry Lamm, Andy C. Y. Li, Gabriel Perdue, Sergei V.
[489] Timothy C. Ralph, Alexei Gilchrist, Gerard J. Milburn, Isakov, Orion Martin, Ross Thomson, and Jackson Beall,
William J. Munro, and Scott Glancy, Quantum computa- et al., in 2021 IEEE/ACM Second International Workshop
tion with optical coherent states, Phys. Rev. A 68, 042319 on Quantum Computing Software (QCS) (IEEE, 2021),
(2003). p. 72.

027001-65
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

[505] Steven Weinberg, Phenomenological lagrangians, Phys. A McClean, Kosuke Mitarai, Xiao Yuan, and Lukasz Cincio,
96, 327 (1979). et al., Variational quantum algorithms, Nat. Rev. Phys. 3,
[506] Enrico Fermi, Tentativo di una teoria dei raggi β, Il Nuovo 625 (2021).
Cimento (1924-1942) 11, 1 (1934). [525] Tzu-Ching Yen, Vladyslav Verteletskyi, and Artur
[507] Steven Weinberg, Nuclear forces from chiral lagrangians, F. Izmaylov, Measuring all compatible operators in
Phys. Lett., (Sect.) B;(Netherlands) 251, 00 (1990). one series of single-qubit measurements using unitary
[508] Mannque Rho, Exchange Currents from Chiral Lagrang- transformations, J. Chem. Theory Comput. 16, 2400
ians, Phys. Rev. Lett. 66, 1275 (1991). (2020).
[509] C. Ordonez, L. Ray, and U. Van Kolck, Two-nucleon [526] Xavier Bonet-Monroig, Ryan Babbush, and Thomas E.
potential from chiral lagrangians, Phys. Rev., C. Nucl. O’Brien, Nearly Optimal Measurement Scheduling for
Phys. 53, 2086 (1996). Partial Tomography of Quantum States, Phys. Rev. X 10,
[510] Steven Weinberg, Three-body interactions among nucle- 031064 (2020).
ons and pions, Phys. Lett. B 295, 114 (1992). [527] William J. Huggins, Jarrod R. McClean, Nicholas C.
[511] Paulo F. Bedaque and Ubirajara Van Kolck, Nucleon- Rubin, Zhang Jiang, Nathan Wiebe, K. Birgitta Whaley,
deuteron scattering from an effective field theory, Phys. and Ryan Babbush, Efficient and noise resilient mea-
Lett. B 428, 221 (1998). surements for quantum chemistry on near-term quantum
[512] David B. Kaplan, Martin J. Savage, and Mark B. Wise, computers, npj Quantum Inf. 7, 1 (2021).
A new expansion for nucleon-nucleon interactions, Phys. [528] Guillermo García-Pérez, Matteo A. C. Rossi, Boris
Lett. B 424, 390 (1998). Sokolov, Francesco Tacchino, Panagiotis Kl Barkoutsos,
[513] Paulo F. Bedaque, H.-W. Hammer, and Ubirajara Van Guglielmo Mazzola, Ivano Tavernelli, and Sabrina Man-
Kolck, Effective theory for neutron-deuteron scattering: iscalco, Learning to Measure: Adaptive Informationally
Energy dependence, Phys. Rev., C. Nucl. Phys. 58, R641 Complete Generalized Measurements for Quantum Algo-
(1998). rithms, PRX Quantum 2, 040342 (2021).
[514] David B. Kaplan, Martin J. Savage, and Mark B. Wise, [529] Heiko Hergert, A guided tour of ab initio nuclear many-
Two-nucleon systems from effective field theory, Nucl. body theory, Front. Phys. , ().
Phys. B 534, 329 (1998). [530] Kanav Setia and James D. Whitfield, Bravyi-Kitaev super-
[515] U. Van Kolck, Effective field theory of short-range forces, fast simulation of electronic structure on a quantum com-
Nucl. Phys. A 645, 273 (1999). puter, J. Chem. Phys. 148, 164104 (2018).
[516] Michael C. Birse, Judith A. McGovern, and Keith G. [531] Mark Steudtner and Stephanie Wehner, Fermion-to-qubit
Richardson, A renormalisation-group treatment of two- mappings with varying resource requirements for quantum
body scattering, Phys. Lett. B 464, 169 (1999). simulation, New J. Phys. 20, 063010 (2018).
[517] Simon Williams, Sarah Malik, Michael Spannowsky, and [532] Earl Campbell, Random Compiler for Fast Hamiltonian
Khadeejah Bepari, A quantum walk approach to simulat- Simulation, Phys. Rev. Lett. 123, 070503 (2019).
ing parton showers, (2021), ArXiv:2109.13975. [533] Alessandro Roggero, Spectral-density estimation with the
[518] Sebastian Macaluso and Kyle Cranmer, The quantum Gaussian integral transform, Phys. Rev. A 102, 022409
trellis: A classical algorithm for sampling the parton (2020).
shower with interference effects, (2021), arXiv preprint [534] Heinrich Leutwyler, On the foundations of chiral pertur-
ArXiv:2112.12795. bation theory, Ann. Phys. (N. Y) 235, 165 (1994).
[519] Vincenzo Cirigliano, Zohreh Davoudi, Wouter Dekens, [535] Jürg Gasser and Heinrich Leutwyler, Chiral perturbation
Jordy de Vries, Jonathan Engel, Xu Feng, Julia Gehrlein, theory to one loop, Ann. Phys. (N. Y) 158, 142 (1984).
and Michael Graesser, et al., Neutrinoless double-beta [536] Stefan Scherer, in Advances in Nuclear Physics, Volume
decay: A roadmap for matching theory to experiment, 27 (Springer, 2003), p. 277.
(2022) ArXiv:2203.12169. [537] Veronique Bernard, Chiral perturbation theory and baryon
[520] Ruprecht Machleidt and David Rodrignez Entem, Chiral properties, Prog. Part. Nucl. Phys. 60, 82 (2008).
effective field theory and nuclear forces, Phys. Rep. 503, 1 [538] Thomas Becher and Heinrich Leutwyler, Baryon chiral
(2011). perturbation theory in manifestly Lorentz invariant form,
[521] Bela Bauer, Sergey Bravyi, Mario Motta, and Garnet Kin- Eur. Phys. J. C-Part. Fields 9, 643 (1999).
Lic Chan, Quantum algorithms for quantum chemistry [539] Elizabeth Jenkins and Aneesh V. Manohar, Baryon chi-
and quantum materials science, Chem. Rev. 120, 12685 ral perturbation theory using a heavy fermion lagrangian,
(2020). Phys. Lett. B 255, 558 (1991).
[522] Sam McArdle, Suguru Endo, Alán Aspuru-Guzik, Simon [540] John B. Kogut and Mikhail A. Stephanov, The Phases
C. Benjamin, and Xiao Yuan, Quantum computational of Quantum Chromodynamics: From Confinement to
chemistry, Rev. Mod. Phys. 92, 015003 (2020). Extreme Environments Vol. 21 (Cambridge University
[523] Yudong Cao, Jonathan Romero, Jonathan P. Olson, Press, 2003).
Matthias Degroote, Peter D. Johnson, Mária Kieferová, [541] Walter Hofstetter and Tao Qin, Quantum simulation of
Ian D. Kivlichan, Tim Menke, Borja Peropadre, and Nico- strongly correlated condensed matter systems, J. Phys. B:
las P. D. Sawaya, et al., Quantum chemistry in the age of At., Mol. Opt. Phys. 51, 082001 (2018).
quantum computing, Chem. Rev. 119, 10856 (2019). [542] Tim Langen, Remi Geiger, and Jörg Schmiedmayer, Ultra-
[524] Marco Cerezo, Andrew Arrasmith, Ryan Babbush, Simon cold atoms out of equilibrium, Annu. Rev. Condens.
C. Benjamin, Suguru Endo, Keisuke Fujii, Jarrod R. Matter Phys. 6, 201 (2015).

027001-66
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

[543] Cheng Chin, Rudolf Grimm, Paul Julienne, and Eite [555] J. Zhang, G. Pagano, P. W. Hess, A. Kyprianidis, P.
Tiesinga, Feshbach resonances in ultracold gases, Rev. Becker, H. Kaplan, A. V. Gorshkov, Z. X. Gong, and C.
Mod. Phys. 82, 1225 (2010). Monroe, Observation of a many-body dynamical phase
[544] Xiaoling Wu, Xinhui Liang, Yaoqi Tian, Fan Yang, Cheng transition with a 53-qubit quantum simulator, Nature 551,
Chen, Yong-Chun Liu, Meng Khoon Tey, and Li You, A 601 (2017).
concise review of Rydberg atom based quantum compu- [556] Petar Jurcevic, Ben P. Lanyon, Philipp Hauke, Cor-
tation and quantum simulation, Chin. Phys. B 30, 020305 nelius Hempel, Peter Zoller, Rainer Blatt, and Christian F.
(2021). Roos, Quasiparticle engineering and entanglement propa-
[545] Manuel Endres, Hannes Bernien, Alexander Keesling, gation in a quantum many-body system, Nature 511, 202
Harry Levine, Eric R. Anschuetz, Alexandre Krajen- (2014).
brink, Crystal Senko, Vladan Vuletic, Markus Greiner, [557] Philip Richerme, Zhe-Xuan Gong, Aaron Lee, Crystal
and Mikhail D. Lukin, Atom-by-atom assembly of defect- Senko, Jacob Smith, Michael Foss-Feig, Spyridon Micha-
free one-dimensional cold atom arrays, Science 354, 1024 lakis, Alexey V. Gorshkov, and Christopher Monroe, Non-
(2016). local propagation of correlations in quantum systems with
[546] Ivaylo S. Madjarov, Jacob P. Covey, Adam L. Shaw, long-range interactions, Nature 511, 198 (2014).
Joonhee Choi, Anant Kale, Alexandre Cooper, Hannes [558] P. Jurcevic, H. Shen, P. Hauke, C. Maier, T. Brydges, C.
Pichler, Vladimir Schkolnik, Jason R. Williams, and Hempel, B. P. Lanyon, M. Heyl, R. Blatt, and C. F. Roos,
Manuel Endres, High-fidelity entanglement and detec- Direct Observation of Dynamical Quantum Phase Tran-
tion of alkaline-earth Rydberg atoms, Nat. Phys. 16, 857 sitions in an Interacting Many-Body System, Phys. Rev.
(2020). Lett. 119, 080501 (2017).
[547] Harry Levine, Alexander Keesling, Ahmed Omran, [559] W. L. Tan, P. Becker, F. Liu, G. Pagano, K. S. Collins, A.
Hannes Bernien, Sylvain Schwartz, Alexander S. Zibrov, De, L. Feng, H. B. Kaplan, A. Kyprianidis, R. Lundgren,
Manuel Endres, Markus Greiner, Vladan Vuletić, and W. Morong, S. Whitsitt, A. V. Gorshkov, and C. Mon-
Mikhail D. Lukin, High-Fidelity Control and Entangle- roe, Domain-wall confinement and dynamics in a quantum
ment of Rydberg-Atom Qubits, Phys. Rev. Lett. 121, simulator, Nat. Phys. 17, 742 (2021).
123603 (2018). [560] D. Porras and J. I. Cirac, Effective Quantum Spin Systems
[548] Harry Levine, Alexander Keesling, Giulia Semeghini, with Trapped Ions, Phys. Rev. Lett. 92, 207901 (2004).
Ahmed Omran, Tout T. Wang, Sepehr Ebadi, Hannes [561] Lucas Lamata, Antonio Mezzacapo, Jorge Casanova,
Bernien, Markus Greiner, Vladan Vuletić, and Hannes and Enrique Solano, Efficient quantum simulation of
Pichler, et al., Parallel Implementation of High-Fidelity fermionic and bosonic models in trapped ions, EPJ Quan-
Multiqubit Gates with Neutral Atoms, Phys. Rev. Lett. tum Technol. 1, 1 (2014).
123, 170503 (2019). [562] A. Mezzacapo, J. Casanova, L. Lamata, and E. Solano,
[549] Dolev Bluvstein, Harry Levine, Giulia Semeghini, Tout Digital Quantum Simulation of the Holstein Model in
T. Wang, Sepehr Ebadi, Marcin Kalinowski, Alexander Trapped Ions, Phys. Rev. Lett. 109, 200501 (2012).
Keesling, Nishad Maskara, Hannes Pichler, and Markus [563] Lincoln D. Carr, David DeMille, Roman V. Krems, and
Greiner, et al., A quantum processor based on coherent Jun Ye, Cold and ultracold molecules: Science, technol-
transport of entangled atom arrays, (2021), arXiv preprint ogy and applications, New J. Phys. 11, 055049 (2009).
ArXiv:2112.03923. [564] Bryce Gadway and Bo Yan, Strongly interacting ultra-
[550] Hannes Pichler, Sheng-Tao Wang, Leo Zhou, Soonwon cold polar molecules, J. Phys. B: At., Mol. Opt. Phys. 49,
Choi, and Mikhail D. Lukin, Quantum optimization for 152002 (2016).
maximum independent set using Rydberg atom arrays, [565] Lee R. Liu, Jonathan D. Hood, Yichao Yu, Jessie T.
(2018), arXiv preprint ArXiv:1808.10816. Zhang, Kenneth Wang, Y.-W. Lin, Till Rosenband, and
[551] P. Scholl, H. J. Williams, G. Bornet, F. Wallner, D. K.-K. Ni, Molecular Assembly of Ground-State Cooled
Barredo, T. Lahaye, A. Browaeys, L. Henriet, A. Sig- Single Atoms, Phys. Rev. X 9, 021039 (2019).
noles, and C. Hainaut, et al., Microwave-engineering of [566] Loïc Anderegg, Lawrence W. Cheuk, Yicheng Bao,
programmable XXZ hamiltonians in arrays of Rydberg Sean Burchesky, Wolfgang Ketterle, Kang-Kuen Ni, and
atoms, (2021), arXiv preprint ArXiv:2107.14459. John M. Doyle, An optical tweezer array of ultracold
[552] Mark Saffman, Thad G. Walker, and Klaus Mølmer, Quan- molecules, Science 365, 1156 (2019).
tum information with Rydberg atoms, Rev. Mod. Phys. 82, [567] Varun D. Vaidya, Yudan Guo, Ronen M. Kroeze, Kyle E.
2313 (2010). Ballantine, Alicia J. Kollár, Jonathan Keeling, and Ben-
[553] Mohammadsadegh Khazali and Klaus Mølmer, Fast jamin L. Lev, Tunable-Range, Photon-Mediated Atomic
Multiqubit Gates by Adiabatic Evolution in Interacting Interactions in Multimode Cavity QED, Phys. Rev. X 8,
Excited-State Manifolds of Rydberg Atoms and Supercon- 011002 (2018).
ducting Circuits, Phys. Rev. X 10, 021054 (2020). [568] Changsuk Noh and Dimitris G. Angelakis, Quantum sim-
[554] Joseph W. Britton, Brian C. Sawyer, Adam C. Keith, ulations and many-body physics with light, Rep. Prog.
C.-C. Joseph Wang, James K. Freericks, Hermann Uys, Phys. 80, 016401 (2016).
Michael J. Biercuk, and John J. Bollinger, Engineered [569] Michael J. Hartmann, Quantum simulation with interact-
two-dimensional Ising interactions in a trapped-ion quan- ing photons, J. Opt. 18, 104005 (2016).
tum simulator with hundreds of spins, Nature 484, 489 [570] Stephan Welte, Bastian Hacker, Severin Daiss, Stephan
(2012). Ritter, and Gerhard Rempe, Photon-mediated quantum

027001-67
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

gate between two trapped neutral atoms in an optical [584] Abhinav Kandala, Antonio Mezzacapo, Kristan Temme,
cavity, (2018), arXiv preprint ArXiv:1801.05980. Maika Takita, Markus Brink, Jerry M. Chow, and Jay M.
[571] Alexandre Blais, Arne L. Grimsmo, S. M. Girvin, Gambetta, Hardware-efficient variational quantum eigen-
and Andreas Wallraff, Circuit quantum electrodynamics, solver for small molecules and quantum magnets, Nature
Rev. Mod. Phys. 93, 025005 (2021). 549, 242 (2017).
[572] Sebastian Schmidt and Jens Koch, Circuit QED lattices: [585] Eric T. Holland, Kyle A. Wendt, Konstantinos Kravvaris,
Towards quantum simulation with superconducting cir- Xian Wu, W. Erich Ormand, Jonathan L. DuBois, Sofia
cuits, Ann. Phys. 525, 395 (2013). Quaglioni, and Francesco Pederiva, Optimal control for
[573] Xiqiao Wang, Ehsan Khatami, Fan Fei, Jonathan Wyrick, the quantum simulation of nuclear dynamics, Phys. Rev.
Pradeep Namboodiri, Ranjit Kashid, Albert F. Rigosi, A 101, 062307 (2020).
Garnett Bryant, and Richard Silver, Quantum simulation [586] Alexei Kitaev and William A. Webb, Wavefunction prepa-
of an extended Fermi-Hubbard model using a 2D lat- ration and resampling using a quantum computer, (2008),
tice of dopant-based quantum dots, (2021), arXiv preprint arXiv preprint ArXiv:0801.0342.
ArXiv:2110.08982. [587] Sukin Sim, Peter D. Johnson, and Alán Aspuru-Guzik,
[574] Patrick Harvey-Collard, N. Tobias Jacobson, Martin Expressibility and entangling capability of parameterized
Rudolph, Jason Dominguez, Gregory A. Ten Eyck, Joel quantum circuits for hybrid quantum-classical algorithms,
R Wendt, Tammy Pluym, John King Gamble, Michael Adv. Quantum Technol. 2, 1900070 (2019).
P. Lilly, and Michel Pioro-Ladrière, et al., Coherent cou- [588] Lena Funcke, Tobias Hartung, Karl Jansen, Stefan Kühn,
pling between a quantum dot and a donor in silicon, Nat. and Paolo Stornati, Dimensional expressivity analy-
Commun. 8, 1 (2017). sis of parametric quantum circuits, Quantum 5, 422
[575] Yu He, S. K. Gorman, Daniel Keith, Ludwik Kranz, J. G. (2021).
Keizer, and M. Y. Simmons, A two-qubit gate between [589] Tianyi Peng, Aram W. Harrow, Maris Ozols, and Xiaodi
phosphorus donor electrons in silicon, Nature 571, 371 Wu, Simulating Large Quantum Circuits on a Small Quan-
(2019). tum Computer, Phys. Rev. Lett. 125, 150504 (2020).
[576] Jonathan Wyrick, Xiqiao Wang, Ranjit V. Kashid, Pradeep [590] Chao Song, Kai Xu, Wuxin Liu, Chui-ping Yang, Shi-
Namboodiri, Scott W. Schmucker, Joseph A. Hagmann, Biao Zheng, Hui Deng, Qiwei Xie, Keqiang Huang, Qiu-
Keyi Liu, Michael D. Stewart Jr, Curt A. Richter, and Gar- jiang Guo, and Libo Zhang, et al., 10-Qubit Entanglement
nett W. Bryant, et al., Atom-by-atom fabrication of single and Parallel Logic Operations with a Superconducting
and few dopant quantum devices, Adv. Funct. Mater. 29, Circuit, Phys. Rev. Lett. 119, 180511 (2017).
1903475 (2019). [591] S. Y. Yanzhu Chen, Maziar Farahzad, and T. C. Wei,
[577] Xiqiao Wang, Jonathan Wyrick, Ranjit V. Kashid, Pradeep Detector tomography on IBM 5-qubit quantum computers
Namboodiri, Scott W. Schmucker, Andrew Murphy, and mitigation of imperfect measurement, (2019), arXiv
Michael D. Stewart, and Richard M. Silver, Atomic-scale preprint ArXiv:1904.11935.
control of tunneling in donor-based devices, Commun. [592] Michael R. Geller and Mingyu Sun, Efficient correction
Phys. 3, 1 (2020). of multiqubit measurement errors, (2020), arXiv preprint
[578] Andrew W. Cross, Lev S. Bishop, Sarah Sheldon, Paul D. ArXiv:2001.09980.
Nation, and Jay M. Gambetta, Validating quantum com- [593] Benjamin Nachman and Michael R. Geller, Categorizing
puters using randomized model circuits, Phys. Rev. A 100, readout error correlations on near term quantum comput-
032328 (2019). ers, (2021), arXiv preprint ArXiv:2104.04607.
[579] IonQ Staff, “Algorithmic Qubits: A Better Single-Number [594] Ken X. Wei, Isaac Lauer, Srikanth Srinivasan, Neereja
Metric,” [Link] Sundaresan, Douglas T. McClure, David Toyli, David C.
thmic-qubits. McKay, Jay M. Gambetta, and Sarah Sheldon, Verify-
[580] Kenneth R. Brown, Jungsang Kim, and Christopher Mon- ing multipartite entangled Greenberger-Horne-Zeilinger
roe, Co-designing a scalable quantum computer with states via multiple quantum coherences, Phys. Rev. A 101,
trapped atomic ions, npj Quantum Inf. 2, 16034 (2016). 032343 (2020).
[581] D. Kielpinski, C. Monroe, and D. J. Wineland, Architec- [595] Kathleen E. Hamilton, Tyler Kharazi, Titus Morris,
ture for a large-scale ion-trap quantum computer, Nature Alexander J. McCaskey, Ryan S. Bennink, and Raphael C.
417, 709 (2002). Pooser, in 2020 IEEE International Conference on Quan-
[582] Laird Egan, Dripto M. Debroy, Crystal Noel, Andrew tum Computing and Engineering (QCE) (IEEE, 2020),
Risinger, Daiwei Zhu, Debopriyo Biswas, Michael New- p. 430.
man, Muyuan Li, Kenneth R. Brown, Marko Cetina, and [596] Benjamin Nachman, Miroslav Urbanek, Wibe A. de Jong,
Christopher Monroe, Fault-tolerant control of an error- and Christian W. Bauer, Unfolding quantum computer
corrected qubit, Nature 598, 281 (2021). readout noise, npj Quantum Inf. 6, 1 (2020).
[583] Lukas Postler, Sascha Heussen, Ivan Pogorelov, Manuel [597] Jakob M. Günther, Francesco Tacchino, James R. Woot-
Rispler, Thomas Feldker, Michael Meth, Christian ton, Ivano Tavernelli, and Panagiotis Kl Barkoutsos,
D. Marciniak, Roman Stricker, Martin Ringbauer, Improving readout in quantum simulations with repetition
Rainer Blatt, Philipp Schindler, Markus Muller, and codes, Quantum Sci. Technol. 7, 015009 (2021).
Thomas Monz, Demonstration of fault-tolerant univer- [598] Rebecca Hicks, Bryce Kobrin, Christian W. Bauer, and
sal quantum gate operations, (2021), arXiv preprint Benjamin Nachman, Active readout-error mitigation,
ArXiv:2111.12654. Phys. Rev. A 105, 012419 (2022).

027001-68
QUANTUM SIMULATION FOR HIGH-ENERGY PHYSICS PRX QUANTUM 4, 027001 (2023)

[599] Abhinav Kandala, Kristan Temme, Antonio D. Córcoles, A retargetable compiler for NISQ devices, Quantum Sci.
Antonio Mezzacapo, Jerry M. Chow, and Jay M. Gam- Technol. 6, 014003 (2020).
betta, Error mitigation extends the computational reach of [613] Ed. Younis, Koushik Sen, Katherine Yelick, and Costin
a noisy quantum processor, Nature 567, 491 (2019). Iancu, Qfast: in 2021 IEEE International Conference
[600] Lena Funcke, Tobias Hartung, Karl Jansen, Stefan Kühn, on Quantum Computing and Engineering (QCE) (IEEE,
Paolo Stornati, and Xiaoyang Wang, Measurement error 2021), p. 232.
mitigation in quantum computers through classical bit-flip [614] John van de Wetering, ZX-calculus for the work-
correction, (2020), arXiv preprint ArXiv:2007.03663. ing quantum computer scientist, (2020), arXiv preprint
[601] Paul D. Nation, Hwajung Kang, Neereja Sundaresan, and ArXiv:2012.13966.
Jay M. Gambetta, Scalable Mitigation of Measurement [615] Suguru Endo, Zhenyu Cai, Simon C. Benjamin, and Xiao
Errors on Quantum Computers, PRX Quantum 2, 040326 Yuan, Hybrid quantum-classical algorithms and quan-
(2021). tum error mitigation, J. Phys. Soc. Jpn. 90, 032001
[602] Alex Mott, Joshua Job, Jean Roch Vlimant, Daniel Lidar, (2021).
and Maria Spiropulu, Solving a Higgs optimization prob- [616] William J. Huggins, Sam McArdle, Thomas E. O’Brien,
lem with quantum annealing for machine learning, Nature Joonho Lee, Nicholas C. Rubin, Sergio Boixo, K. Birgitta
550, 375 (2017). Whaley, Ryan Babbush, and Jarrod R. McClean, Virtual
[603] Alexander Zlokapa, Abhishek Anand, Jean-Roch Vli- Distillation for Quantum Error Mitigation, Phys. Rev. X
mant, Javier M. Duarte, Joshua Job, Daniel Lidar, and 11, 041036 (2021).
Maria Spiropulu, Charged particle tracking with quantum [617] Jarrod R. McClean, Nicholas C. Rubin, Kevin J. Sung,
annealing-inspired optimization, Quantum Machine Intell. Ian D. Kivlichan, Xavier Bonet-Monroig, Yudong Cao,
3, 27 (2021). ArXiv:1908.04475, Chengyu Dai, E. Schuyler Fried, Craig Gidney, and Bren-
[604] Frédéric Bapst, Wahid Bhimji, Paolo Calafiura, Heather dan Gimby, et al., Openfermion: The electronic structure
Gray, Wim Lavrijsen, Lucy Linder, and Alex Smith, package for quantum computers, Quantum Sci. Technol.
A pattern recognition algorithm for quantum annealers, 5, 034014 (2020).
Comput. Softw. Big Sci. 4, 1 (2020). [618] Ville Bergholm, Josh Izaac, Maria Schuld, Christian
[605] Erik Gustafson, Patrick Dreher, Zheyue Hang, and Yan- Gogolin, M. Sohaib Alam, Shahnawaz Ahmed, Juan
nick Meurice, Benchmarking quantum computers for Miguel Arrazola, Carsten Blank, Alain Delgado, and
real-time evolution of a (1 + 1) field theory with error Soran Jahangiri, et al., Pennylane: Automatic differenti-
mitigation, Quantum Sci. Technol. 6, 045020 (2021). ation of hybrid quantum-classical computations, (2018),
ArXiv:1910.09478. arXiv preprint ArXiv:1811.04968.
[606] Alexander Erhard, Joel J. Wallman, Lukas Postler, [619] See a Snowmass whitepaper on “Quantum Computing
Michael Meth, Roman Stricker, Esteban A. Martinez, Systems and Software for High-energy Physics Research”
Philipp Schindler, Thomas Monz, Joseph Emerson, and [635] for further discussions.
Rainer Blatt, Characterizing large-scale quantum com- [620] “XACC: A System-level Software Infrastructure for Het-
puters via cycle benchmarking, Nat. Commun. 10, 1 erogeneous Quantum-Classical Computing,” [Link]
(2019). [Link]/highlight/xacc-system-level-software-infrastruct
[607] Kubra Yeter-Aydeniz, Zachary Parks, Aadithya Nair, Erik ure-heterogeneous-quantum-classical-computing.
Gustafson, Alexander F. Kemper, Raphael C. Pooser, [621] Thien Nguyen, Lindsay Bassman, Phillip C. Lotshaw,
Yannick Meurice, and Patrick Dreher, Measuring NISQ Dmitry Lyakh, Alexander McCaskey, Vicente Leyton-
gate-based qubit stability using a 1+1 D field theory and Ortega, Raphael Pooser, Wael Elwasif, Travis S. Humble,
cycle benchmarking, (2022), arXiv preprint ArXiv:2201. and Wibe A. de Jong, Quasimo: A composable library
02899. to program hybrid workflows for quantum simulation,
[608] M. D. Anis, H. Abraham, R. A. AduOffei, G. Agliardi, M. (2021), arXiv preprint ArXiv:2105.07993.
Aharoni, I. Y. Akhalwaya, G. Aleksandrowicz, T. Alexan- [622] D. V. Babukhin, A. A. Zhukov, and W. V. Pogosov,
der, M. Amy, and S. Anagolum, Qiskit: An open-source Hybrid digital-analog simulation of many-body dynamics
framework for quantum computing, (2021),. with superconducting qubits, Phys. Rev. A 101, 052337
[609] Nathan Killoran, Josh Izaac, Nicolás Quesada, Ville (2020).
Bergholm, Matthew Amy, and Christian Weedbrook, [623] Tasio Gonzalez-Raya, Rodrigo Asensio-Perea, Ana Mar-
Strawberry fields: A software platform for photonic quan- tin, Lucas C. Céleri, Mikel Sanz, Pavel Lougovski, and
tum computing, Quantum 3, 129 (2019). Eugene F. Dumitrescu, Digital-Analog Quantum Simula-
[610] Cirq Developers, “Cirq,” (2021), See full list of authors on tions using the Cross-Resonance Effect, PRX Quantum 2,
Github: [Link] 020328 (2021).
butors. [624] Peter A. Boyle, Dong Chen, Norman H. Christ, Michael
[611] Krysta Svore, Alan Geller, Matthias Troyer, John Azariah, A. Clark, Saul D. Cohen, Calin Cristian, Zhihua Dong,
Christopher Granade, Bettina Heim, Vadym Kliuchnikov, Alan Gara, Balint Joo, and Chulwoo Jung, et al., Overview
Mariia Mykhailova, Andres Paz, and Martin Roetteler, in of the QCDSP and QCDOC computers, IBM J. Res. Dev.
Proceedings of the real world domain specific languages 49, 351 (2005).
workshop 2018 (2018), p. 1. [625] Peter A. Boyle, Dong Chen, Norman H. Christ, Mike
[612] Seyon Sivarajah, Silas Dilkes, Alexander Cowtan, Will Clark, Saul Cohen, Zhihua Dong, Alan Gara, Balint
Simmons, Alec Edgington, and Ross Duncan, t|ket >: Joo, Chulwoo Jung, and Ludmila Levkova, et al., in

027001-69
CHRISTIAN W. BAUER et al. PRX QUANTUM 4, 027001 (2023)

SC’04: Proceedings of the 2004 ACM/IEEE Conference [631] “Expanding capacity quantum information science and
on Supercomputing (IEEE, 2004), p. 40. engineering,” [Link]
[626] Mathieu Pollet, “France opens access to quantum comput- expandqise.
ing to researchers, start-ups,” [Link] [632] “Scientific Discovery Through Advanced Computing,”
section / digital / news / france - opens - access - to - quantum- [Link]
computing-to-researchers-start-ups/. [633] “The InQubator for Quantum Simulation workshop: At
[627] C. Gougard, “Germany unveils powerful quantum the Interface of Quantum Sensors and Quantum Simula-
computer to keep Europe in global tech race,” https:// tions,” [Link]
[Link]/article/germany-unveils-europes-first- [634] Yannick Meurice, James C. Osborn, Ryo Sakai, Judah
quantum-computer/. Unmuth-Yockey, Simon Catterall, and Rolando D.
[628] K. Flinders, “The UK government is backing a quantum Somma, Tensor networks for high energy physics: Con-
computing project that will provide businesses with access tribution to Snowmass 2021, (2022), arXiv preprint
to the technology,” [Link] ArXiv:2203.04902.
news/252488484/UK-to-build-commercially-available- [635] Travis S. Humble, Andrea Delgado, Raphael Pooser,
quantum-computer. Christopher Seck, Ryan Bennink, Vicente Leyton-Ortega,
[629] National Academies of Sciences, Engineering, and C.-C. Joseph Wang, Eugene Dumitrescu, Titus Mor-
Medicine: Quantum Computing: Progress and Prospects, ris, and Kathleen Hamilton, et al., Snowmass white
edited by Emily Grumbling and Mark Horowitz (The paper: Quantum computing systems and software for
National Academies Press, Washington, DC, 2019). high-energy physics research, (2022), arXiv preprint
[630] “QuarkNet program,” [Link] ArXiv:2203.07091.

027001-70