Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916219
M. Harris, M. H. Langston, Pierre-David Létourneau, G. Papanicolaou, J. Ezick, R. Lethin
We present a fast, large-scale algorithm for the simulation of electromagnetic waves (Maxwell’s equations) in three-dimensional inhomogeneous media. The algorithm has a complexity of $O(Nlog (N))$ and runs in parallel. Numerical simulations show the rapid treatment of problems with tens of millions of unknowns on a small shared-memory cluster (≤ 16 cores).
{"title":"Fast Large-Scale Algorithm for Electromagnetic Wave Propagation in 3D Media","authors":"M. Harris, M. H. Langston, Pierre-David Létourneau, G. Papanicolaou, J. Ezick, R. Lethin","doi":"10.1109/HPEC.2019.8916219","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916219","url":null,"abstract":"We present a fast, large-scale algorithm for the simulation of electromagnetic waves (Maxwell’s equations) in three-dimensional inhomogeneous media. The algorithm has a complexity of $O(Nlog (N))$ and runs in parallel. Numerical simulations show the rapid treatment of problems with tens of millions of unknowns on a small shared-memory cluster (≤ 16 cores).","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124772697","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916436
Mohamad Sindi, John R. Williams
We share experiences in implementing a containerbased HPC environment that could help sustain running HPC workloads on clusters. By running workloads inside containers, we are able to migrate them from cluster nodes anticipating hardware problems, to healthy nodes while the workloads are running. Migration is done using the CRIU tool with no application modification. No major interruption or overhead is introduced to the workload. Various real HPC applications are tested. Tests are done with different hardware node specs, network interconnects, and MPI implementations. We also benchmark the applications on containers and compare performance to native. Results demonstrate successful migration of HPC workloads inside containers with minimal interruption, while maintaining the integrity of the results produced. We provide several YouTube videos demonstrating the migration tests. Benchmarks also show that application performance on containers is close to native. We discuss some of the challenges faced during implementation and solutions adopted. To the best of our knowledge, we believe this work is the first to demonstrate successful migration of real MPI-based HPC workloads using CRIU and containers.
{"title":"Using Container Migration for HPC Workloads Resilience","authors":"Mohamad Sindi, John R. Williams","doi":"10.1109/HPEC.2019.8916436","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916436","url":null,"abstract":"We share experiences in implementing a containerbased HPC environment that could help sustain running HPC workloads on clusters. By running workloads inside containers, we are able to migrate them from cluster nodes anticipating hardware problems, to healthy nodes while the workloads are running. Migration is done using the CRIU tool with no application modification. No major interruption or overhead is introduced to the workload. Various real HPC applications are tested. Tests are done with different hardware node specs, network interconnects, and MPI implementations. We also benchmark the applications on containers and compare performance to native. Results demonstrate successful migration of HPC workloads inside containers with minimal interruption, while maintaining the integrity of the results produced. We provide several YouTube videos demonstrating the migration tests. Benchmarks also show that application performance on containers is close to native. We discuss some of the challenges faced during implementation and solutions adopted. To the best of our knowledge, we believe this work is the first to demonstrate successful migration of real MPI-based HPC workloads using CRIU and containers.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128319792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916446
M. H. Langston, M. Harris, Pierre-David Létourneau, R. Lethin, J. Ezick
The Combinatorial Multigrid (CMG) technique is a practical and adaptable solver and combinatorial preconditioner for solving certain classes of large, sparse systems of linear equations. CMG is similar to Algebraic Multigrid (AMG) but replaces large groupings of fine-level variables with a single coarse-level one, resulting in simple and fast interpolation schemes. These schemes further provide control over the refinement strategies at different levels of the solver hierarchy depending on the condition number of the system being solved [1]. While many pre-existing solvers may be able to solve large, sparse systems with relatively low complexity, inversion may require O(n2) space; whereas, if we know that a linear operator has $tilde{n}=O(n)$ nonzero elements, we desire to use O(n) space in order to reduce communication as much as possible. Being able to invert sparse linear systems of equations, asymptotically as fast as the values can be read from memory, has been identified by the Defense Advanced Research Projects Agency (DARPA) and the Department of Energy (DOE) as increasingly necessary for scalable solvers and energy-efficient algorithms [2], [3] in scientific computing. Further, as industry and government agencies move towards exascale, fast solvers and communication-avoidance will be more necessary [4], [5]. In this paper, we present an optimized implementation of the Combinatorial Multigrid in C using Petsc and analyze the solution of various systems using the CMG approach as a preconditioner on much larger problems than have been presented thus far. We compare the number of iterations, setup times and solution times against other popular preconditioners for such systems, including Incomplete Cholesky and a Multigrid approach in Petsc against common problems, further exhibiting superior performance by the CMG.1 2
{"title":"Combinatorial Multigrid: Advanced Preconditioners For Ill-Conditioned Linear Systems","authors":"M. H. Langston, M. Harris, Pierre-David Létourneau, R. Lethin, J. Ezick","doi":"10.1109/HPEC.2019.8916446","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916446","url":null,"abstract":"The Combinatorial Multigrid (CMG) technique is a practical and adaptable solver and combinatorial preconditioner for solving certain classes of large, sparse systems of linear equations. CMG is similar to Algebraic Multigrid (AMG) but replaces large groupings of fine-level variables with a single coarse-level one, resulting in simple and fast interpolation schemes. These schemes further provide control over the refinement strategies at different levels of the solver hierarchy depending on the condition number of the system being solved [1]. While many pre-existing solvers may be able to solve large, sparse systems with relatively low complexity, inversion may require O(n2) space; whereas, if we know that a linear operator has $tilde{n}=O(n)$ nonzero elements, we desire to use O(n) space in order to reduce communication as much as possible. Being able to invert sparse linear systems of equations, asymptotically as fast as the values can be read from memory, has been identified by the Defense Advanced Research Projects Agency (DARPA) and the Department of Energy (DOE) as increasingly necessary for scalable solvers and energy-efficient algorithms [2], [3] in scientific computing. Further, as industry and government agencies move towards exascale, fast solvers and communication-avoidance will be more necessary [4], [5]. In this paper, we present an optimized implementation of the Combinatorial Multigrid in C using Petsc and analyze the solution of various systems using the CMG approach as a preconditioner on much larger problems than have been presented thus far. We compare the number of iterations, setup times and solution times against other popular preconditioners for such systems, including Incomplete Cholesky and a Multigrid approach in Petsc against common problems, further exhibiting superior performance by the CMG.1 2","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"134 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125475653","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916504
Majid Rasouli, Vidhi Zala, R. Kirby, H. Sundar
Multigrid is one of the most effective methods for solving elliptic PDEs. It is algorithmically optimal and is robust when combined with Krylov methods. Algebraic multigrid is especially attractive due to its blackbox nature. This however comes at the cost of increased setup costs that can be significant in case of systems where the system matrix changes frequently making it difficult to amortize the setup cost. In this work, we investigate several strategies for performing lazy updates to the multigrid hierarchy corresponding to changes in the system matrix. These include delayed updates, value updates without changing structure, process local changes, and full updates. We demonstrate that in many cases, the overhead of building the AMG hierarchy can be mitigated for rapidly changing system matrices.
{"title":"Scalable Lazy-update Multigrid Preconditioners","authors":"Majid Rasouli, Vidhi Zala, R. Kirby, H. Sundar","doi":"10.1109/HPEC.2019.8916504","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916504","url":null,"abstract":"Multigrid is one of the most effective methods for solving elliptic PDEs. It is algorithmically optimal and is robust when combined with Krylov methods. Algebraic multigrid is especially attractive due to its blackbox nature. This however comes at the cost of increased setup costs that can be significant in case of systems where the system matrix changes frequently making it difficult to amortize the setup cost. In this work, we investigate several strategies for performing lazy updates to the multigrid hierarchy corresponding to changes in the system matrix. These include delayed updates, value updates without changing structure, process local changes, and full updates. We demonstrate that in many cases, the overhead of building the AMG hierarchy can be mitigated for rapidly changing system matrices.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120894534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916495
M. Rogowski, Suha N. Kayum
Load balancing is a crucial factor affecting the performance of parallel applications. Improper work distribution leads to underutilization of computing resources and an unnecessary increase in runtime. This paper identifies the imbalance sources in reservoir simulation and characterizes them as static or dynamic. Simulation model properties that change over time, such as well management actions, are registered and correlated with performance characteristics hence identifying sources of imbalance. The results are exploratory and used to validate the current approach of static grid-to-process, and well-to-process assignment widely used in commercial parallel reservoir simulators. Areas in which implementing dynamic load balancing would be worthwhile are identified.
{"title":"Evaluation of the Imbalance Evolution in Parallel Reservoir Simulation","authors":"M. Rogowski, Suha N. Kayum","doi":"10.1109/HPEC.2019.8916495","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916495","url":null,"abstract":"Load balancing is a crucial factor affecting the performance of parallel applications. Improper work distribution leads to underutilization of computing resources and an unnecessary increase in runtime. This paper identifies the imbalance sources in reservoir simulation and characterizes them as static or dynamic. Simulation model properties that change over time, such as well management actions, are registered and correlated with performance characteristics hence identifying sources of imbalance. The results are exploratory and used to validate the current approach of static grid-to-process, and well-to-process assignment widely used in commercial parallel reservoir simulators. Areas in which implementing dynamic load balancing would be worthwhile are identified.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121806033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916328
Hongkuan Zhou, Ajitesh Srivastava, R. Kannan, V. Prasanna
PageRank is a fundamental graph algorithm to evaluate the importance of vertices in a graph. In this paper, we present an efficient parallel PageRank design based on an edge-centric scatter-gather model. To overcome the poor locality of PageRank and optimize the memory performance, we develop a fast and efficient partitioning technique. We first partition all the vertices into non-overlapping vertex sets such that the data of each vertex set can fit in the cache; then we sort the outgoing edges of each vertex set based on the destination vertices to minimize random memory writes. The partitioning technique significantly reduces random accesses to main memory and improves the sustained memory bandwidth by 3×. It also enables efficient parallel execution on multicore platforms; we use distinct cores to execute the computations of distinct vertex sets in parallel to achieve speedup. We implement our design on a 16-core Intel Xeon processor and use various large-scale real-life and synthetic datasets for evaluation. Compared with the PageRank Pipeline Benchmark, our design achieves 12× to 19× speedup for all the datasets.
运行时可重新配置的软件与可重新配置的硬件相结合是非常可取的，因为这是在不损害可编程性的情况下最大化运行时效率的一种手段。这类软件系统的编译器设计起来极其困难，因为它们必须在运行时利用不同类型的硬件。为了解决与动态可重构硬件相匹配的工作流的静态和动态编译器优化的需要，我们提出了一种针对软件定义硬件的动态软件编译器的中心组件的新设计。我们的综合设计不仅关注静态知识，还关注从程序执行中提取知识的半监督式提取，并开发其性能模型。具体来说，我们的新动态和可扩展知识库1)在工作流执行期间持续收集知识2)在最佳(可用)硬件配置上确定工作流的最佳实现。它在存储来自编译器的其他组件以及人工分析人员的信息并向其提供信息方面起着中心作用。通过丰富的三部分图表示，知识库捕获并学习了有关分解和将代码步骤映射到内核以及将内核映射到可用硬件配置的广泛信息。该知识库使用$ c++ $ Boost库实现，能够快速处理离线和在线查询和更新。我们展示了我们的知识库可以在$1 ms$内回答查询，而不管它存储的工作流的数量。据我们所知，这是支持高级语言编译以利用任意可重构平台的第一个动态和可扩展知识库的设计。
{"title":"Design and Implementation of Knowledge Base for Runtime Management of Software Defined Hardware","authors":"Hongkuan Zhou, Ajitesh Srivastava, R. Kannan, V. Prasanna","doi":"10.1109/HPEC.2019.8916328","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916328","url":null,"abstract":"PageRank is a fundamental graph algorithm to evaluate the importance of vertices in a graph. In this paper, we present an efficient parallel PageRank design based on an edge-centric scatter-gather model. To overcome the poor locality of PageRank and optimize the memory performance, we develop a fast and efficient partitioning technique. We first partition all the vertices into non-overlapping vertex sets such that the data of each vertex set can fit in the cache; then we sort the outgoing edges of each vertex set based on the destination vertices to minimize random memory writes. The partitioning technique significantly reduces random accesses to main memory and improves the sustained memory bandwidth by 3×. It also enables efficient parallel execution on multicore platforms; we use distinct cores to execute the computations of distinct vertex sets in parallel to achieve speedup. We implement our design on a 16-core Intel Xeon processor and use various large-scale real-life and synthetic datasets for evaluation. Compared with the PageRank Pipeline Benchmark, our design achieves 12× to 19× speedup for all the datasets.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131738217","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916301
A. Gjersvik, Robert J. Moss
With a rapidly growing and evolving National Airspace System (NAS), ACAS X is intended to be the nextgeneration airborne collision avoidance system that can meet the demands its predecessor could not. The ACAS X algorithms are developed in the Julia programming language and are exercised in simulation environments tailored to test different characteristics of the system. Massive parallelization of these simulation environments has been implemented on the Lincoln Laboratory Supercomputing Center cluster in order to expedite the design and performance optimization of the system. This work outlines the approach to parallelization of one of our simulation tools and presents the resulting simulation speedups as well as a discussion on how it will enhance system characterization and design. Parallelization has made our simulation environment 33 times faster, which has greatly sped up the development process of ACAS X.
{"title":"A Parallel Simulation Approach to ACAS X Development","authors":"A. Gjersvik, Robert J. Moss","doi":"10.1109/HPEC.2019.8916301","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916301","url":null,"abstract":"With a rapidly growing and evolving National Airspace System (NAS), ACAS X is intended to be the nextgeneration airborne collision avoidance system that can meet the demands its predecessor could not. The ACAS X algorithms are developed in the Julia programming language and are exercised in simulation environments tailored to test different characteristics of the system. Massive parallelization of these simulation environments has been implemented on the Lincoln Laboratory Supercomputing Center cluster in order to expedite the design and performance optimization of the system. This work outlines the approach to parallelization of one of our simulation tools and presents the resulting simulation speedups as well as a discussion on how it will enhance system characterization and design. Parallelization has made our simulation environment 33 times faster, which has greatly sped up the development process of ACAS X.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133462714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916215
Xu T. Liu, J. Firoz, Marcin Zalewski, M. Halappanavar, K. Barker, A. Lumsdaine, A. Gebremedhin
Designing a scalable algorithm for community detection is challenging due to the simultaneous need for both high performance and quality of solution. We propose a new distributed algorithm for community detection based on a novel Label Propagation algorithm. The algorithm is inspired by the direction optimization technique in graph traversal algorithms, relies on the use of frontiers, and alternates between abstractions called label push and label pull. This organization creates flexibility and affords us with opportunities for balancing performance and quality of solution. We implement our algorithm in distributed memory with the active-message based asynchronous many-task runtime AM++. We experiment with two seeding strategies for the initial seeding stage, namely, random seeding and degree seeding. With the Graph Challenge dataset, our distributed implementation, in conjunction with the runtime support, detects the communities in graphs having 20 million vertices in less than one second while achieving reasonably high quality of solution.
{"title":"Distributed Direction-Optimizing Label Propagation for Community Detection","authors":"Xu T. Liu, J. Firoz, Marcin Zalewski, M. Halappanavar, K. Barker, A. Lumsdaine, A. Gebremedhin","doi":"10.1109/HPEC.2019.8916215","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916215","url":null,"abstract":"Designing a scalable algorithm for community detection is challenging due to the simultaneous need for both high performance and quality of solution. We propose a new distributed algorithm for community detection based on a novel Label Propagation algorithm. The algorithm is inspired by the direction optimization technique in graph traversal algorithms, relies on the use of frontiers, and alternates between abstractions called label push and label pull. This organization creates flexibility and affords us with opportunities for balancing performance and quality of solution. We implement our algorithm in distributed memory with the active-message based asynchronous many-task runtime AM++. We experiment with two seeding strategies for the initial seeding stage, namely, random seeding and degree seeding. With the Graph Challenge dataset, our distributed implementation, in conjunction with the runtime support, detects the communities in graphs having 20 million vertices in less than one second while achieving reasonably high quality of solution.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132792541","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916476
Dominik Tödling, Martin Winter, M. Steinberger
Breadth-First Search is an important basis for many different graph-based algorithms with applications ranging from peer-to-peer networking to garbage collection. However, the performance of different approaches depends strongly on the type of graph. In this paper, we present an efficient algorithm that performs well on a variety of different graphs. As part of this, we look into utilizing dynamic parallelism in order to both reduce overhead from latency between the CPU and GPU, as well as speed up the algorithm itself. Lastly, integrate the algorithm with the faimGraph framework for dynamic graphs and examine the relative performance to a Compressed-Sparse-Row data structure. We show that our algorithm can be well adapted to the dynamic setting and outperforms another competing dynamic graph framework on our test set.
{"title":"Breadth-First Search on Dynamic Graphs using Dynamic Parallelism on the GPU","authors":"Dominik Tödling, Martin Winter, M. Steinberger","doi":"10.1109/HPEC.2019.8916476","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916476","url":null,"abstract":"Breadth-First Search is an important basis for many different graph-based algorithms with applications ranging from peer-to-peer networking to garbage collection. However, the performance of different approaches depends strongly on the type of graph. In this paper, we present an efficient algorithm that performs well on a variety of different graphs. As part of this, we look into utilizing dynamic parallelism in order to both reduce overhead from latency between the CPU and GPU, as well as speed up the algorithm itself. Lastly, integrate the algorithm with the faimGraph framework for dynamic graphs and examine the relative performance to a Compressed-Sparse-Row data structure. We show that our algorithm can be well adapted to the dynamic setting and outperforms another competing dynamic graph framework on our test set.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"34 5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117278166","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916550
T. Davis, M. Aznaveh, Scott P. Kolodziej
SuiteSparse:GraphBLAS is a full implementation of the GraphBLAS standard, which provides a powerful and expressive framework for creating graph algorithms based on the elegant mathematics of sparse matrix operations on a semiring. Algorithms written in GraphBLAS achieve high performance with minimal development time. Using GraphBLAS, it took a mere 20 minutes to write a first-cut computational kernel that solves the Sparse Deep Neural Network Graph Challenge. Understanding the problem description and file format, writing code to read in the files that define the problem, and comparing our results with the reference solution took a full day. The kernel consists of a single for-loop around 4 lines of code, all of which are calls to GraphBLAS, and it worked perfectly the first time it was compiled. The sequential performance of the GraphBLAS solution is 3x to 5x faster than the MATLAB reference implementation. OpenMP parallelism gives an additional 10x to 15x speedup on a 20-core Intel processor, 17x on an IBM Power8 system, and 20x on a Power9 system, for the largest problems. Since SuiteSparse:GraphBLAS does not yet employ MPI, this was added at the application level, a development effort that took one week, primarily because of difficulties in resolving a load-balancing issue in the MPI-based parallel algorithm.
{"title":"Write Quick, Run Fast: Sparse Deep Neural Network in 20 Minutes of Development Time via SuiteSparse:GraphBLAS","authors":"T. Davis, M. Aznaveh, Scott P. Kolodziej","doi":"10.1109/HPEC.2019.8916550","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916550","url":null,"abstract":"SuiteSparse:GraphBLAS is a full implementation of the GraphBLAS standard, which provides a powerful and expressive framework for creating graph algorithms based on the elegant mathematics of sparse matrix operations on a semiring. Algorithms written in GraphBLAS achieve high performance with minimal development time. Using GraphBLAS, it took a mere 20 minutes to write a first-cut computational kernel that solves the Sparse Deep Neural Network Graph Challenge. Understanding the problem description and file format, writing code to read in the files that define the problem, and comparing our results with the reference solution took a full day. The kernel consists of a single for-loop around 4 lines of code, all of which are calls to GraphBLAS, and it worked perfectly the first time it was compiled. The sequential performance of the GraphBLAS solution is 3x to 5x faster than the MATLAB reference implementation. OpenMP parallelism gives an additional 10x to 15x speedup on a 20-core Intel processor, 17x on an IBM Power8 system, and 20x on a Power9 system, for the largest problems. Since SuiteSparse:GraphBLAS does not yet employ MPI, this was added at the application level, a development effort that took one week, primarily because of difficulties in resolving a load-balancing issue in the MPI-based parallel algorithm.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115479197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：