Parallel sparse linear solver with GMRES method using minimization techniques of communications for GPU clusters

被引：0
|
作者：
Lilia Ziane Khodja
Raphaël Couturier
Arnaud Giersch
Jacques M. Bahi
机构：
[1] University of Franche-Comte,FEMTO
[2] IUT Belfort-Montbéliard,ST Institute
来源：
The Journal of Supercomputing | 2014年 / 69卷
关键词：
Parallel GMRES; Cluster of GPUs; Communication reduction;
D O I：
暂无
中图分类号：
学科分类号：
摘要：
In this paper, we aim at exploiting the power computing of a graphics processing unit (GPU) cluster for solving large sparse linear systems. We implement the parallel algorithm of the generalized minimal residual iterative method using the Compute Unified Device Architecture programming language and the MPI parallel environment. The experiments show that a GPU cluster is more efficient than a CPU cluster. In order to optimize the performances, we use a compressed storage format for the sparse vectors and the hypergraph partitioning. These solutions improve the spatial and temporal localization of the shared data between the computing nodes of the GPU cluster.
引用
收藏
页码：200 / 224
页数：24
相关论文
共 50 条
  • [1] Parallel sparse linear solver with GMRES method using minimization techniques of communications for GPU clusters
    Khodja, Lilia Ziane
    Couturier, Raphael
    Giersch, Arnaud
    Bahi, Jacques M.
    JOURNAL OF SUPERCOMPUTING, 2014, 69 (01): : 200 - 224
  • [2] Parallel Sparse Linear Solver GMRES for GPU Clusters with Compression of Exchanged Data
    Bahi, Jacques M.
    Couturier, Raphael
    Khodja, Lilia Ziane
    EURO-PAR 2011: PARALLEL PROCESSING WORKSHOPS, PT I, 2012, 7155 : 471 - 480
  • [3] Parallel GMRES implementation for solving sparse linear systems on GPU clusters
    Bahi, Jacques M.
    Couturier, Raphael
    Khodja, Lilia Ziane
    HIGH PERFORMANCE COMPUTING SYMPOSIUM 2011 (HPC 2011) - 2011 SPRING SIMULATION MULTICONFERENCE - BK 6 OF 8, 2011, 43 (02): : 12 - 19
  • [4] Fast GMRES-GPU solver for large scale sparse linear systems
    Liu, Youquan
    Yin, Kangxue
    Wu, Enhua
    Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics, 2011, 23 (04): : 553 - 560
  • [5] Parallel GMRES solver for fast analysis of large linear dynamic systems on GPU platforms
    He, Kai
    Tan, Sheldon X. -D
    Zhao, Hengyang
    Liu, Xue-Xin
    Wang, Hai
    Shi, Guoyong
    INTEGRATION-THE VLSI JOURNAL, 2016, 52 : 10 - 22
  • [6] An Integer Arithmetic-Based Sparse Linear Solver Using a GMRES Method and Iterative Refinement
    Iwashita, Takeshi
    Suzuki, Kengo
    Fukaya, Takeshi
    PROCEEDINGS OF SCALA 2020: 11TH WORKSHOP ON LATEST ADVANCES IN SCALABLE ALGORITHMS FOR LARGE-SCALE SYSTEMS, 2020, : 1 - 8
  • [7] Parallel Power Grid Analysis Using Preconditioned GMRES Solver on CPU-GPU Platforms
    Liu, Xue-Xin
    Wang, Hai
    Tan, Sheldon X-D
    2013 IEEE/ACM INTERNATIONAL CONFERENCE ON COMPUTER-AIDED DESIGN (ICCAD), 2013, : 561 - 568
  • [8] Parallel Structured Sparse Triangular Solver for GPU Platform
    Chen D.-K.
    Yang C.
    Liu F.-F.
    Ma W.-J.
    Ruan Jian Xue Bao/Journal of Software, 2023, 34 (11): : 4941 - 4951
  • [9] Concurrent Number Cruncher:: An efficient sparse linear solver on the GPU
    Buatois, Luc
    Caumon, Guillaume
    Levy, Bruno
    HIGH PERFORMANCE COMPUTING AND COMMUNICATIONS, PROCEEDINGS, 2007, 4782 : 358 - +
  • [10] PSPIKE: A Parallel Hybrid Sparse Linear System Solver
    Manguoglu, Murat
    Sameh, Ahmed H.
    Schenk, Olaf
    EURO-PAR 2009: PARALLEL PROCESSING, PROCEEDINGS, 2009, 5704 : 797 - +
12345