计算机科学 ›› 2015, Vol. 42 ›› Issue (11): 63-64.doi: 10.11896/j.issn.1002-137X.2015.11.012

• 2014年全国高性能计算机学术年会 • 上一篇    下一篇

面向定制结构的稀疏矩阵分块方法

邬贵明,王 淼,谢向辉,窦 勇,郭 松   

  1. 国防科学技术大学计算机学院 长沙410073;数学工程与先进计算国家重点实验室 无锡214125,数学工程与先进计算国家重点实验室 无锡214125,数学工程与先进计算国家重点实验室 无锡214125,国防科学技术大学计算机学院 长沙410073,国防科学技术大学计算机学院 长沙410073
  • 出版日期:2018-11-14 发布日期:2018-11-14
  • 基金资助:
    本文受国家自然科学基金项目(91430214),湖南省科研资助

Sparse Matrix Blocking Method for Custom Architecture

WU Gui-ming, WANG Miao, XIE Xiang-hui, DOU Yong and GUO Song   

  • Online:2018-11-14 Published:2018-11-14

摘要: 稀疏矩阵向量乘是科学计算的核心问题,采用定制结构来加速稀疏矩阵向量乘的执行对提升科学计算性能具有重要意义。针对目前面向定制结构的稀疏矩阵分块方法和表示方法的缺点,提出了稀疏矩阵二维均匀分块方法和相应的表示方法嵌套分块CSR。实验结果表明,提出的稀疏矩阵分块方法和表示方法能够有效减少填零个数。

关键词: 稀疏矩阵向量乘,定制结构,稀疏矩阵,数据分块

Abstract: Sparse matrix vector multiplication is one of the most important applications in scientific computing.Using custom architectures to implement sparse matrix vector multiplication is introduced to improve the performance of scientific computing.To address the problem in existing sparse matrix blocking method and representation,a two-dimension uniform blocking method and its according representation were proposed in this paper.The experimental results show that the proposed sparse matrix blocking method and representation can reduce the padding zero significantly.

Key words: Sparse matrix vector multiplication,Custom architecture,Sparse matrix,Data blocking

[1] Dorrance R,Ren F,Markovi′ D.A Scalable Sparse Matrix-Vector Multiplication Kernel for Energy-Efficient Sparse-BLAS on FPGAs[C]∥Proceedings of the 2014 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays (FPGA).ACM,2014:161-170
[2] Fowers J,Ovtcharov K,Strauss K,et al.A High Memory Bandwidth FPGA Accelerator for Sparse Matrix-Vector Multiplication[C]∥Proceedings of the 2014 IEEE 22nd Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM).IEEE,2014:36-43
[3] Zhuo L,Prasanna V K.Sparse Matrix-Vector Multiplication on FPGAs[C]∥Proceedings of the 13th ACM/SIGDA International Symposium on Field Programmable Gate Arrays (FPGA).ACM,2005:63-74
[4] Sun J,Peterson G D,Storaasli O O.Sparse Matrix-Vector Multiplication Design on FPGAs[C]∥ Proceedings of the 15th AnnualIEEE Symposium Field-Programmable Custom Computing Machines (FCCM).IEEE,2007:349-352
[5] Stathis P T.Sparse Matrix Vector Processing Formats[D].Delft,Netherlands:Delft University of Technology,2004
[6] Williams S,Vuduc R,Oliker L,et al.Optimizing Sparse Matrix-Vector Multiply on Emerging Multicore Platforms[J].Journal of Parallel Computing,2009,35(3):178-194
[7] The University of Florida Sparse Matrix Collection.http://www.cise.ufl.edu/research/sparse/matrices

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!