Asymmetric multicore processors (AMPs) have recently emerged as an appealing technology for sever... more Asymmetric multicore processors (AMPs) have recently emerged as an appealing technology for severely energy-constrained environments, especially in mobile appliances where heterogeneity in applications is mainstream. In addition, given the growing interest for low-power high performance computing, this type of architectures is also being investigated as a means to improve the throughput-per-Watt of complex scientific applications. In this paper, we design and embed several architecture-aware optimizations into a multi-threaded general matrix multiplication (gemm), a key operation of the BLAS, in order to obtain a high performance implementation for ARM big.LITTLE AMPs. Our solution is based on the reference implementation of gemm in the BLIS library, and integrates a cache-aware configuration as well as asymmetric--static and dynamic scheduling strategies that carefully tune and distribute the operation's micro-kernels among the big and LITTLE cores of the target processor. The ...
Proceedings of the 3rd IEEE/ACM International Conference on Big Data Computing, Applications and Technologies
Hybrid cloud bursting (i.e., leasing temporary off-premise cloud resources to boost the capacity ... more Hybrid cloud bursting (i.e., leasing temporary off-premise cloud resources to boost the capacity during peak utilization), has made significant impact especially for big data analytics, where the explosion of data sizes and increasingly complex computations frequently leads to insufficient local data center capacity. Cloud bursting however introduces a major challenge to runtime systems due to the limited throughput and high latency of data transfers between onpremise and off-premise resources (weak link). This issue and how to address it is not well understood. We contribute with a comprehensive study on what challenges arise in this context, what potential strategies can be applied to address them and what best practices can be leveraged in real-life. Specifically, we focus our study on iterative MapReduce applications, which are a class of large-scale data intensive applications particularly popular on hybrid clouds. In this context, we study how data locality can be leveraged over the weak link both from the storage layer perspective (when and how to move it off-premise) and from the scheduling perspective (when to compute off-premise). We conclude with a brief discussion on how to set up an experimental framework suitable to study the effectiveness of our proposal in future work.
The road towards Exascale Computing requires a holistic effort to address three different challen... more The road towards Exascale Computing requires a holistic effort to address three different challenges simultaneously: high performance, energy efficiency, and programmability. The use of runtime task schedulers to orchestrate parallel executions with minimal developer intervention has been introduced in recent years to tackle the programmability issue while maintaining, or even improving, performance. In this paper, we enhance the SuperMatrix runtime task scheduler integrated in the libflame library in two different directions that address high performance and energy efficiency. First, we extend the runtime by accommodating hybrid parallel executions and managing task priorities for dense linear algebra operations, with remarkable performance improvements. Second, we introduce techniques to reduce energy consumption during idle times inherent to parallel executions, attaining important energy savings. In addition, we propose a power consumption model that can be leveraged by runtime ...
With the appearance of multi-/many core machines, applications and runtime systems evolved in ord... more With the appearance of multi-/many core machines, applications and runtime systems evolved in order to exploit the new on-node concurrency brought by new software paradigms. POSIX threads (Pthreads) was widely-adopted for that purpose and it remains as the most used threading solution in current hardware. Lightweight thread (LWT) libraries emerged as an alternative offering lighter mechanisms to tackle the massive concurrency of current hardware. In this paper, we analyze in detail the most representative threading libraries including Pthread-and LWT-based solutions. In addition, to examine the suitability of LWTs for different use cases, we develop a set of microbenchmarks consisting of OpenMP patterns commonly found in current parallel codes, and we compare the results using threading libraries and OpenMP implementations. Moreover, we study the semantics offered by threading libraries in order to expose the similarities among different LWT application programming interfaces and their advantages over Pthreads. This study exposes that LWT libraries outperform solutions based on operating system threads when tasks and nested parallelism are required.
In this article, we put forward a new methodology and strategy for teaching the Instruction Set A... more In this article, we put forward a new methodology and strategy for teaching the Instruction Set Architecture in a "Computer Organization" unit. This unit belongs to the second year of the undergraduate program in Computer Science at our University. In particular, we have centered our effort on the development of laboratory sessions, focused on an adequate introduction of assembler language of a general purpose processor. Our methodology has taken into account, among other aspects, the choosing of processor and tools, the pace with which concepts are introduced and the responsibility of learning. Thus, we obtain a less traumatic approach for our students to one of the most important subjects in the background of our future computer engineers and scientists. Results obtained show students response is positive. This effect is reflected in student's interest and the ease with which they are been able to solve the exercises we set them to do. Because of that, an improvement on learning in this subject and this aspect is reflected in the subsequent evaluation of students.
2011 18th International Conference on High Performance Computing, 2011
The hardware and software advances of Graphics Processing Units (GPUs) have favored the developme... more The hardware and software advances of Graphics Processing Units (GPUs) have favored the development of GPGPU (General-Purpose Computation on GPUs) and its adoption in many scientific, engineering, and industrial areas. Thus, GPUs are increasingly being introduced in high-performance computing systems as well as in datacenters. On the other hand, virtualization technologies are also receiving rising interest in these domains, because of their many benefits on acquisition and maintenance savings. There are currently several works on GPU virtualization. However, there is no standard solution allowing access to GPGPU capabilities from virtual machine environments like, e.g., VMware, Xen, VirtualBox, or KVM. Such lack of a standard solution is delaying the integration of GPGPU into these domains. In this paper, we propose a first step towards a general and open source approach for using GPGPU features within VMs. In particular, we describe the use of rCUDA, a GPGPU (General-Purpose Computation on GPUs) virtualization framework, to permit the execution of GPU-accelerated applications within virtual machines (VMs), thus enabling GPGPU capabilities on any virtualized environment. Our experiments with rCUDA in the context of KVM and VirtualBox on a system equipped with two NVIDIA GeForce 9800 GX2 cards illustrate the overhead introduced by the rCUDA middleware and prove the feasibility and scalability of this general virtualizing solution. Experimental results show that the overhead is proportional to the dataset size, while the scalability is similar to that of the native environment.
The International Journal of High Performance Computing Applications, 2012
The following article has been retracted from publication in International Journal of High Perfor... more The following article has been retracted from publication in International Journal of High Performance Computing Applications due to duplicate publication arising from error. Color and texture analysis on emerging parallel architectures Francisco D Igual, Rafael Mayo, Timothy DR Hartley, Ümit V Çatalyürek, Antonio Ruiz, and Manuel Ujaldon, International Journal of High Performance Computing Applications August 2012 26:237-259. Below version of the article stands as the version on record. Color and texture analysis using emerging parallel architectures Francisco D Igual, Rafael Mayo, Timothy DR Hartley, Ümit V Çatalyürek, Antonio Ruiz, and Manuel Ujaldon, International Journal of High Performance Computing Applications November 2011 25: 404-427. Our special thanks to the author for drawing attention to the duplication, thereby allowing for the literature to be corrected.
International Journal of High Performance Computing Applications, 2011
While image texture is effective for use in pattern-recognition and image-analysis algorithms, te... more While image texture is effective for use in pattern-recognition and image-analysis algorithms, textural features are time-consuming to calculate on standard CPUs. Therefore, we present novel implementations of textural-feature algorithms on graphics processors (GPUs), enabling fast color and texture analysis. Since different textural-feature calculations exhibit diverse characteristics, we focus on using general and algorithm-specific techniques to exploit the inherent parallelism and computational power of a GPU. Common operations required during the textural-feature pipeline range from streaming computations to recursive procedures, from arithmetically intensive transcendental functions to matrix operations. Some of these kernels are well-suited to GPUs, while others require considerable programming effort to fully exploit the memory hierarchy due to their memory-usage patterns. In this paper, different strategies for computing textural features on GPUs are compared with counterpa...
We present several algorithms to compute the solution of a linear system of equations on a GPU, a... more We present several algorithms to compute the solution of a linear system of equations on a GPU, as well as general techniques to improve their performance, such as padding and hybrid GPU-CPU computation. We also show how iterative refinement with mixed-precision can be used to regain full accuracy in the solution of linear systems. Experimental results on a G80 using
The increase in performance of the last generations of graphics processors (GPUs) has made this c... more The increase in performance of the last generations of graphics processors (GPUs) has made this class of hardware a coprocessing platform of remarkable success in certain types of operations. In this paper we evaluate the performance of linear algebra and image processing routines, both on classical and unified GPU architectures and traditional processors (CPUs). From this study, we gain insights on the properties that make an algorithm likely to deliver high perfor-mance on a GPU. Keywords: Graphics processors (GPUs), general purpose computing on GPU, linear algebra, image processing, high performance. 1 Departamento de Ingenieŕıa y Ciencia de los Computadores
$ ./gb_run.sh Writing down cpu and gpu info... Copying LaTeX templates to gpubenchmark_tesla2_res... more $ ./gb_run.sh Writing down cpu and gpu info... Copying LaTeX templates to gpubenchmark_tesla2_results/templates/... Copying 'gb_make_pdf' to gpubenchmark_tesla2_results/... Running transfer speed test between main memory and hard disk: [********] Running transfer speed tests between main memory and GPU... Running transfer speed test between main memory and GPU (without padding): [********] Running transfer speed test between main memory and GPU (with padding): [********] Running transfer speed test between two gpus: [********] Running matrix per matrix tests... Running matrix per matrix test (1/4): [******] Running matrix per matrix test (2/4): [******] Running matrix per matrix test (3/4): [******] Running matrix per matrix test (4/4): [******] Running matrix per vector tests... Running matrix per vector test (1/4): [********] Running matrix per vector test (2/4): [********] Running matrix per vector test (3/4): [********] Running matrix per vector test (4/4): [********] The GPUBenchmark tests have ended. You can now either run the script: './gpubenchmark_tesla2_results/gb_make_pdf.sh'; or copy the contents of the './gpubenchmark_tesla2_results/' directory to another computer and run the 'gb_make_pdf.sh' script there. para distintas máquinas pueden consultarse en la siguiente dirección: http://lorca.act.uji.es/projects/gpubenchmark/ GPUBenchmark results for tesla2
Large-scale distributed systems (e.g., datacenters, HPC systems, clouds, large-scale networks, et... more Large-scale distributed systems (e.g., datacenters, HPC systems, clouds, large-scale networks, etc.) consume and will consume enormous amounts of energy. Therefore, accurately monitoring the power dissipation and energy consumption of these systems is more unavoidable. The main novelty of this contribution is the analysis and evaluation of different external and internal power monitoring devices tested using two different computing systems, a server and a desktop machine. Furthermore, we provide experimental results for a variety of benchmarks which intensively exercise the main components (CPU, Memory, HDDs, and NICs) of the target platforms to validate the accuracy of the equipment in terms of power dissipation and energy consumption. On the other hand, we also evaluate three different power measurement interfaces available on current architecture generations. Thanks to the high sampling rate and to the different measured lines, the internal wattmeters allow an improved visualizat...
Many current high-performance clusters include one or more GPUs per node in order to dramatically... more Many current high-performance clusters include one or more GPUs per node in order to dramatically reduce application execution time, but the utilization of these accelerators is usually far below 100%. In this context, remote GPU virtualization can help to reduce acquisition costs as well as the overall energy consumption. In this paper, we investigate the potential overhead and bottlenecks of several “heterogeneous” scenarios consisting of client GPU-less nodes running CUDA applications and remote GPUequipped server nodes providing access to NVIDIA hardware accelerators. The experimental evaluation is performed using three general-purpose multicore processors (Intel Xeon, Intel Atom and ARM Cortex A9), two graphics accelerators (NVIDIA GeForce GTX480 and NVIDIA Quadro M1000), and two relevant scientific applications (CUDASW++ and LAMMPS) arising in bioinformatics and molecular dynamics simulations.
This paper addresses the efficient explotation of task-level parallelism, present in many dense l... more This paper addresses the efficient explotation of task-level parallelism, present in many dense linear algebra operations, from the point of view of both computational performance and energy consumption. In particular, we consider a procedure, the Slack Reduction Algorithm (SRA), to optimize the execution frequency of a collection of tasks (in which many dense linear algebra algorithms can be decomposed) on multicore architectures. The results from this procedure are modulated by an energy-aware simulator, which is in charge of scheduling/mapping the execution of these tasks to the cores, leveraging dynamic frequency voltage scaling featured by current technology. Simultaneously, the simulator evaluates the performance benefits of the solution. Experiments with these tools show significant energy gains for two key dense linear algebra operations: the Cholesky and QR factorizations.
Asymmetric processors have emerged as an appealing technology for severely energy-constrained env... more Asymmetric processors have emerged as an appealing technology for severely energy-constrained environments, especially in the mobile market where heterogeneity in applications is mainstream. In addition, given the growing interest on ultra low-power architectures for high performance computing, this type of platforms are also being investigated in the road towards the implementation of energy- efficient high-performance scientific applications. In this paper, we propose a first step towards a complete implementation of the BLAS interface adapted to asymmetric ARM big.LITTLE processors, analyzing the trade-offs between performance and energy efficiency when compared to existing homogeneous (symmetric) multi-threaded BLAS implementations. Our experimental results reveal important gains in performance while maintaining the energy efficiency of homogeneous solutions by efficiently exploiting all the resources of the asymmetric processor.
We describe a prototype web service for model reduction of very large-scale linear systems, with ... more We describe a prototype web service for model reduction of very large-scale linear systems, with dimension in the order of millions of states, that includes a user-friendly interface designed so that the computation can be easily performed via the HTTP protocol. Access via a web browser isolates the user of the service from the complexities of installing and using the parallel model reduction codes and the maintenance of the hardware. In case the routines are found to be appropriate for the problem the user needs to solve, the library can be then downloaded and installed on the user's own computing resources. This paper illustrates the major issues of the access procedure by means of graphical examples, and describes the structure and implementation of the remote model reduction service. The service is offered in a cluster of Linux machines. 1. Introduction. The development of the Internet and its popularization via the world wide web (www) is producing a tremendous impact in ou...
We propose two high-level application programming interfaces (APIs) to use a graphics processing ... more We propose two high-level application programming interfaces (APIs) to use a graphics processing unit (GPU) as a coprocessor for dense linear algebra operations. Combined with an extension of the FLAME API and an implementation on top of NVIDIA CUBLAS, the result is an efficient and user-friendly tool to design, implement, and execute dense linear algebra operations on the current generation of NVIDIA graphics processors, of wide-appeal to scientists and engineers. As an application of the developed APIs, we implement and evaluate the performance of three different variants of the Cholesky factorization.
In recent years, several lightweight thread (LWT) libraries have emerged to tackle exascale chall... more In recent years, several lightweight thread (LWT) libraries have emerged to tackle exascale challenges. These offer programming models (PMs) based on user-level threads and incorporate their own lightweight mechanisms. However, each library proposes its own PM, exposing different semantics and hindering portability. To address this drawback, we have designed Generic Lightweight Thread (GLT), an application programming interface that frames the functionality of the most popular LWT libraries for high-performance computing under a single PM. We implement GLT on top of Argobots, Mas-siveThreads, and Qthreads. We provide GLT as a dynamic library, as well as in the form of a static version based on macro preprocessing resolution to reduce overhead. This paper discusses the GLT PM and demonstrates its minimal performance impact.
Asymmetric multicore processors (AMPs) have recently emerged as an appealing technology for sever... more Asymmetric multicore processors (AMPs) have recently emerged as an appealing technology for severely energy-constrained environments, especially in mobile appliances where heterogeneity in applications is mainstream. In addition, given the growing interest for low-power high performance computing, this type of architectures is also being investigated as a means to improve the throughput-per-Watt of complex scientific applications. In this paper, we design and embed several architecture-aware optimizations into a multi-threaded general matrix multiplication (gemm), a key operation of the BLAS, in order to obtain a high performance implementation for ARM big.LITTLE AMPs. Our solution is based on the reference implementation of gemm in the BLIS library, and integrates a cache-aware configuration as well as asymmetric--static and dynamic scheduling strategies that carefully tune and distribute the operation's micro-kernels among the big and LITTLE cores of the target processor. The ...
Proceedings of the 3rd IEEE/ACM International Conference on Big Data Computing, Applications and Technologies
Hybrid cloud bursting (i.e., leasing temporary off-premise cloud resources to boost the capacity ... more Hybrid cloud bursting (i.e., leasing temporary off-premise cloud resources to boost the capacity during peak utilization), has made significant impact especially for big data analytics, where the explosion of data sizes and increasingly complex computations frequently leads to insufficient local data center capacity. Cloud bursting however introduces a major challenge to runtime systems due to the limited throughput and high latency of data transfers between onpremise and off-premise resources (weak link). This issue and how to address it is not well understood. We contribute with a comprehensive study on what challenges arise in this context, what potential strategies can be applied to address them and what best practices can be leveraged in real-life. Specifically, we focus our study on iterative MapReduce applications, which are a class of large-scale data intensive applications particularly popular on hybrid clouds. In this context, we study how data locality can be leveraged over the weak link both from the storage layer perspective (when and how to move it off-premise) and from the scheduling perspective (when to compute off-premise). We conclude with a brief discussion on how to set up an experimental framework suitable to study the effectiveness of our proposal in future work.
The road towards Exascale Computing requires a holistic effort to address three different challen... more The road towards Exascale Computing requires a holistic effort to address three different challenges simultaneously: high performance, energy efficiency, and programmability. The use of runtime task schedulers to orchestrate parallel executions with minimal developer intervention has been introduced in recent years to tackle the programmability issue while maintaining, or even improving, performance. In this paper, we enhance the SuperMatrix runtime task scheduler integrated in the libflame library in two different directions that address high performance and energy efficiency. First, we extend the runtime by accommodating hybrid parallel executions and managing task priorities for dense linear algebra operations, with remarkable performance improvements. Second, we introduce techniques to reduce energy consumption during idle times inherent to parallel executions, attaining important energy savings. In addition, we propose a power consumption model that can be leveraged by runtime ...
With the appearance of multi-/many core machines, applications and runtime systems evolved in ord... more With the appearance of multi-/many core machines, applications and runtime systems evolved in order to exploit the new on-node concurrency brought by new software paradigms. POSIX threads (Pthreads) was widely-adopted for that purpose and it remains as the most used threading solution in current hardware. Lightweight thread (LWT) libraries emerged as an alternative offering lighter mechanisms to tackle the massive concurrency of current hardware. In this paper, we analyze in detail the most representative threading libraries including Pthread-and LWT-based solutions. In addition, to examine the suitability of LWTs for different use cases, we develop a set of microbenchmarks consisting of OpenMP patterns commonly found in current parallel codes, and we compare the results using threading libraries and OpenMP implementations. Moreover, we study the semantics offered by threading libraries in order to expose the similarities among different LWT application programming interfaces and their advantages over Pthreads. This study exposes that LWT libraries outperform solutions based on operating system threads when tasks and nested parallelism are required.
In this article, we put forward a new methodology and strategy for teaching the Instruction Set A... more In this article, we put forward a new methodology and strategy for teaching the Instruction Set Architecture in a "Computer Organization" unit. This unit belongs to the second year of the undergraduate program in Computer Science at our University. In particular, we have centered our effort on the development of laboratory sessions, focused on an adequate introduction of assembler language of a general purpose processor. Our methodology has taken into account, among other aspects, the choosing of processor and tools, the pace with which concepts are introduced and the responsibility of learning. Thus, we obtain a less traumatic approach for our students to one of the most important subjects in the background of our future computer engineers and scientists. Results obtained show students response is positive. This effect is reflected in student's interest and the ease with which they are been able to solve the exercises we set them to do. Because of that, an improvement on learning in this subject and this aspect is reflected in the subsequent evaluation of students.
2011 18th International Conference on High Performance Computing, 2011
The hardware and software advances of Graphics Processing Units (GPUs) have favored the developme... more The hardware and software advances of Graphics Processing Units (GPUs) have favored the development of GPGPU (General-Purpose Computation on GPUs) and its adoption in many scientific, engineering, and industrial areas. Thus, GPUs are increasingly being introduced in high-performance computing systems as well as in datacenters. On the other hand, virtualization technologies are also receiving rising interest in these domains, because of their many benefits on acquisition and maintenance savings. There are currently several works on GPU virtualization. However, there is no standard solution allowing access to GPGPU capabilities from virtual machine environments like, e.g., VMware, Xen, VirtualBox, or KVM. Such lack of a standard solution is delaying the integration of GPGPU into these domains. In this paper, we propose a first step towards a general and open source approach for using GPGPU features within VMs. In particular, we describe the use of rCUDA, a GPGPU (General-Purpose Computation on GPUs) virtualization framework, to permit the execution of GPU-accelerated applications within virtual machines (VMs), thus enabling GPGPU capabilities on any virtualized environment. Our experiments with rCUDA in the context of KVM and VirtualBox on a system equipped with two NVIDIA GeForce 9800 GX2 cards illustrate the overhead introduced by the rCUDA middleware and prove the feasibility and scalability of this general virtualizing solution. Experimental results show that the overhead is proportional to the dataset size, while the scalability is similar to that of the native environment.
The International Journal of High Performance Computing Applications, 2012
The following article has been retracted from publication in International Journal of High Perfor... more The following article has been retracted from publication in International Journal of High Performance Computing Applications due to duplicate publication arising from error. Color and texture analysis on emerging parallel architectures Francisco D Igual, Rafael Mayo, Timothy DR Hartley, Ümit V Çatalyürek, Antonio Ruiz, and Manuel Ujaldon, International Journal of High Performance Computing Applications August 2012 26:237-259. Below version of the article stands as the version on record. Color and texture analysis using emerging parallel architectures Francisco D Igual, Rafael Mayo, Timothy DR Hartley, Ümit V Çatalyürek, Antonio Ruiz, and Manuel Ujaldon, International Journal of High Performance Computing Applications November 2011 25: 404-427. Our special thanks to the author for drawing attention to the duplication, thereby allowing for the literature to be corrected.
International Journal of High Performance Computing Applications, 2011
While image texture is effective for use in pattern-recognition and image-analysis algorithms, te... more While image texture is effective for use in pattern-recognition and image-analysis algorithms, textural features are time-consuming to calculate on standard CPUs. Therefore, we present novel implementations of textural-feature algorithms on graphics processors (GPUs), enabling fast color and texture analysis. Since different textural-feature calculations exhibit diverse characteristics, we focus on using general and algorithm-specific techniques to exploit the inherent parallelism and computational power of a GPU. Common operations required during the textural-feature pipeline range from streaming computations to recursive procedures, from arithmetically intensive transcendental functions to matrix operations. Some of these kernels are well-suited to GPUs, while others require considerable programming effort to fully exploit the memory hierarchy due to their memory-usage patterns. In this paper, different strategies for computing textural features on GPUs are compared with counterpa...
We present several algorithms to compute the solution of a linear system of equations on a GPU, a... more We present several algorithms to compute the solution of a linear system of equations on a GPU, as well as general techniques to improve their performance, such as padding and hybrid GPU-CPU computation. We also show how iterative refinement with mixed-precision can be used to regain full accuracy in the solution of linear systems. Experimental results on a G80 using
The increase in performance of the last generations of graphics processors (GPUs) has made this c... more The increase in performance of the last generations of graphics processors (GPUs) has made this class of hardware a coprocessing platform of remarkable success in certain types of operations. In this paper we evaluate the performance of linear algebra and image processing routines, both on classical and unified GPU architectures and traditional processors (CPUs). From this study, we gain insights on the properties that make an algorithm likely to deliver high perfor-mance on a GPU. Keywords: Graphics processors (GPUs), general purpose computing on GPU, linear algebra, image processing, high performance. 1 Departamento de Ingenieŕıa y Ciencia de los Computadores
$ ./gb_run.sh Writing down cpu and gpu info... Copying LaTeX templates to gpubenchmark_tesla2_res... more $ ./gb_run.sh Writing down cpu and gpu info... Copying LaTeX templates to gpubenchmark_tesla2_results/templates/... Copying 'gb_make_pdf' to gpubenchmark_tesla2_results/... Running transfer speed test between main memory and hard disk: [********] Running transfer speed tests between main memory and GPU... Running transfer speed test between main memory and GPU (without padding): [********] Running transfer speed test between main memory and GPU (with padding): [********] Running transfer speed test between two gpus: [********] Running matrix per matrix tests... Running matrix per matrix test (1/4): [******] Running matrix per matrix test (2/4): [******] Running matrix per matrix test (3/4): [******] Running matrix per matrix test (4/4): [******] Running matrix per vector tests... Running matrix per vector test (1/4): [********] Running matrix per vector test (2/4): [********] Running matrix per vector test (3/4): [********] Running matrix per vector test (4/4): [********] The GPUBenchmark tests have ended. You can now either run the script: './gpubenchmark_tesla2_results/gb_make_pdf.sh'; or copy the contents of the './gpubenchmark_tesla2_results/' directory to another computer and run the 'gb_make_pdf.sh' script there. para distintas máquinas pueden consultarse en la siguiente dirección: http://lorca.act.uji.es/projects/gpubenchmark/ GPUBenchmark results for tesla2
Large-scale distributed systems (e.g., datacenters, HPC systems, clouds, large-scale networks, et... more Large-scale distributed systems (e.g., datacenters, HPC systems, clouds, large-scale networks, etc.) consume and will consume enormous amounts of energy. Therefore, accurately monitoring the power dissipation and energy consumption of these systems is more unavoidable. The main novelty of this contribution is the analysis and evaluation of different external and internal power monitoring devices tested using two different computing systems, a server and a desktop machine. Furthermore, we provide experimental results for a variety of benchmarks which intensively exercise the main components (CPU, Memory, HDDs, and NICs) of the target platforms to validate the accuracy of the equipment in terms of power dissipation and energy consumption. On the other hand, we also evaluate three different power measurement interfaces available on current architecture generations. Thanks to the high sampling rate and to the different measured lines, the internal wattmeters allow an improved visualizat...
Many current high-performance clusters include one or more GPUs per node in order to dramatically... more Many current high-performance clusters include one or more GPUs per node in order to dramatically reduce application execution time, but the utilization of these accelerators is usually far below 100%. In this context, remote GPU virtualization can help to reduce acquisition costs as well as the overall energy consumption. In this paper, we investigate the potential overhead and bottlenecks of several “heterogeneous” scenarios consisting of client GPU-less nodes running CUDA applications and remote GPUequipped server nodes providing access to NVIDIA hardware accelerators. The experimental evaluation is performed using three general-purpose multicore processors (Intel Xeon, Intel Atom and ARM Cortex A9), two graphics accelerators (NVIDIA GeForce GTX480 and NVIDIA Quadro M1000), and two relevant scientific applications (CUDASW++ and LAMMPS) arising in bioinformatics and molecular dynamics simulations.
This paper addresses the efficient explotation of task-level parallelism, present in many dense l... more This paper addresses the efficient explotation of task-level parallelism, present in many dense linear algebra operations, from the point of view of both computational performance and energy consumption. In particular, we consider a procedure, the Slack Reduction Algorithm (SRA), to optimize the execution frequency of a collection of tasks (in which many dense linear algebra algorithms can be decomposed) on multicore architectures. The results from this procedure are modulated by an energy-aware simulator, which is in charge of scheduling/mapping the execution of these tasks to the cores, leveraging dynamic frequency voltage scaling featured by current technology. Simultaneously, the simulator evaluates the performance benefits of the solution. Experiments with these tools show significant energy gains for two key dense linear algebra operations: the Cholesky and QR factorizations.
Asymmetric processors have emerged as an appealing technology for severely energy-constrained env... more Asymmetric processors have emerged as an appealing technology for severely energy-constrained environments, especially in the mobile market where heterogeneity in applications is mainstream. In addition, given the growing interest on ultra low-power architectures for high performance computing, this type of platforms are also being investigated in the road towards the implementation of energy- efficient high-performance scientific applications. In this paper, we propose a first step towards a complete implementation of the BLAS interface adapted to asymmetric ARM big.LITTLE processors, analyzing the trade-offs between performance and energy efficiency when compared to existing homogeneous (symmetric) multi-threaded BLAS implementations. Our experimental results reveal important gains in performance while maintaining the energy efficiency of homogeneous solutions by efficiently exploiting all the resources of the asymmetric processor.
We describe a prototype web service for model reduction of very large-scale linear systems, with ... more We describe a prototype web service for model reduction of very large-scale linear systems, with dimension in the order of millions of states, that includes a user-friendly interface designed so that the computation can be easily performed via the HTTP protocol. Access via a web browser isolates the user of the service from the complexities of installing and using the parallel model reduction codes and the maintenance of the hardware. In case the routines are found to be appropriate for the problem the user needs to solve, the library can be then downloaded and installed on the user's own computing resources. This paper illustrates the major issues of the access procedure by means of graphical examples, and describes the structure and implementation of the remote model reduction service. The service is offered in a cluster of Linux machines. 1. Introduction. The development of the Internet and its popularization via the world wide web (www) is producing a tremendous impact in ou...
We propose two high-level application programming interfaces (APIs) to use a graphics processing ... more We propose two high-level application programming interfaces (APIs) to use a graphics processing unit (GPU) as a coprocessor for dense linear algebra operations. Combined with an extension of the FLAME API and an implementation on top of NVIDIA CUBLAS, the result is an efficient and user-friendly tool to design, implement, and execute dense linear algebra operations on the current generation of NVIDIA graphics processors, of wide-appeal to scientists and engineers. As an application of the developed APIs, we implement and evaluate the performance of three different variants of the Cholesky factorization.
In recent years, several lightweight thread (LWT) libraries have emerged to tackle exascale chall... more In recent years, several lightweight thread (LWT) libraries have emerged to tackle exascale challenges. These offer programming models (PMs) based on user-level threads and incorporate their own lightweight mechanisms. However, each library proposes its own PM, exposing different semantics and hindering portability. To address this drawback, we have designed Generic Lightweight Thread (GLT), an application programming interface that frames the functionality of the most popular LWT libraries for high-performance computing under a single PM. We implement GLT on top of Argobots, Mas-siveThreads, and Qthreads. We provide GLT as a dynamic library, as well as in the form of a static version based on macro preprocessing resolution to reduce overhead. This paper discusses the GLT PM and demonstrates its minimal performance impact.
Uploads
Papers by rafael MAYO