AI Infra M0 — Foundations
Python, mathematical foundations, Transformers, GPUs, and collective communication.
Course Notes
These are my structured notes from computer science and AI courses, with 37 PDFs in total. Each course includes a short description; click a course title to open its PDF.
AI Infrastructure
Engineering foundations of large-model systems: hardware, communication, training, and inference.
AI Infra M1 — CUDA Programming and Kernel Optimization
CUDA execution model, memory hierarchy, operator implementation, and performance optimization.
AI Infra M2 — Distributed Training
Data, tensor, and pipeline parallelism; collective communication, ZeRO, and training systems.
AI Infra M3 — Inference Optimization
KV caching, batching, quantization, speculative decoding, and inference serving.
DeepSeek-V4.1-Flash — Technical Report Analysis
A modular technical analysis based on the official report, model card, and kernels.
Machine Learning and AI
Machine learning, reinforcement learning, language models, vision, and optimization.
Stanford CS229 — Machine Learning
Supervised and unsupervised learning, probabilistic models, and learning theory.
Stanford CS234 — Reinforcement Learning
MDPs, value iteration, policy gradients, exploration, and deep reinforcement learning.
Stanford CS336 — Language Modeling from Scratch
The full language-model stack, from data and Transformers to training and evaluation.
Stanford CS224n — Natural Language Processing
Word representations, sequence models, attention, pretraining, and NLP applications.
Stanford CS231n — Computer Vision
Convolutional networks, visual recognition, detection and segmentation, and generative models.
UC Berkeley CS285 — Deep Reinforcement Learning
Deep RL, model-based RL, offline RL, and multitask learning.
Stanford EE364a — Convex Optimization
Convex sets, convex functions, duality, optimality conditions, and applied modeling.
Systems, Architecture, and Networks
Computer systems, architecture, operating systems, databases, networks, and security.
CMU 15-213 — CSAPP: Computer Systems
Program representation, processors, memory hierarchy, linking, exceptional control flow, and concurrency.
CMU 15-418/15-618 — Parallel Computer Architecture
Parallel hardware, memory consistency, performance models, and parallel programming.
Stanford CS149 — Parallel Computing
Parallel algorithms, GPUs, heterogeneous computing, and performance analysis.
CMU 15-445/645 — Database Systems
Storage, indexing, query execution, concurrency control, and recovery.
CS111 — Operating Systems
Processes and threads, synchronization, virtual memory, file systems, and OS design.
UC Berkeley CS162 — Operating Systems
Systems programming, processes, scheduling, memory, file systems, and concurrency.
UC Berkeley CS61C — Computer Architecture
RISC-V, digital logic, pipelining, caches, and the memory hierarchy.
UIUC CS425 — Distributed Systems
Replication, consistency, consensus, distributed transactions, and fault tolerance.
UIUC ECE220 — Computer Systems and Programming
C programming, pointers, memory, assembly, and system interfaces.
UIUC ECE408 — Applied Parallel Programming
CUDA, parallel patterns, reductions and scans, tiling, and performance optimization.
Stanford CS144 — Computer Networking
Layered networks, TCP/IP, reliable transport, routing, and application protocols.
UC Berkeley CS161 — Computer Security
Memory safety, cryptography, network security, and system security.
MIT 6.172 — Performance Engineering
Performance modeling, cache optimization, parallelization, compilers, and engineering tuning.
Programming and Software
Programming abstractions, modern C++, and software construction.
Stanford CS106B — Programming Abstractions
Recursion, data structures, algorithm design, and C++ abstraction.
Stanford CS106L — Modern C++
Modern C++ types, the STL, templates, RAII, and move semantics.
MIT 6.031 — Software Construction
Specifications, testing, abstract data types, concurrency, and software design.
Algorithms, Theory, and Mathematics
Algorithms, theory of computation, discrete mathematics, and engineering mathematics foundations.
MIT 6.006 — Introduction to Algorithms
Data structures, graph algorithms, dynamic programming, and complexity.
UC Berkeley CS170 — Efficient Algorithms
Algorithm design, NP-completeness, approximation, and randomized algorithms.
MIT 18.03 — Differential Equations
Ordinary differential equations, linear systems, Fourier methods, and Laplace transforms.
MIT 18.05 — Probability and Statistics
Probability, random variables, estimation, hypothesis testing, and Bayesian methods.
MIT 18.06 — Linear Algebra
Linear systems, vector spaces, eigenvalues, SVD, and applications.
MIT 18.100A — Real Analysis
Limits, continuity, calculus, series, and the foundations of metric spaces.
MIT 18.335 — Numerical Methods
Numerical linear algebra, stability, matrix factorizations, and iterative methods.
MIT 18.404 — Theory of Computation
Automata, computability, complexity, and reductions.
UC Berkeley CS70 — Discrete Mathematics and Probability
Proofs, combinatorics, discrete probability, expectation, and stochastic processes.