期刊文献+

几种矩阵乘并行算法的对比分析 被引量:2

Several Kinds of Parallel Algorithm for Matrix Multiplication Comparative Analysis
在线阅读 下载PDF
导出
摘要 描述了DNS、Cannon、Fox、Systolic矩阵乘并行算法的原理,并对其时间复杂度进行了理论分析。通过对并行算法的各项性能参数的对比分析,得到的结论是DNS算法的时间复杂度最好,但加速比、效率和成本不是最优的。Cannon算法和Fox算法的算法思想类似,但是Cannon算法比Fox算法在数据播送上的花费少,因此整体性能较好。Systolic算法是基于流水线技术的并行矩阵乘算法,有较好的综合性能。 Description of the DNS, Cannon, Fox, Systolic parallel algorithm for matrix multiplication principle, and its time complexity is analyzed. The parallel algorithm of the performance parameters of the comparative analysis, the conclusion is that DNS algorithm time complexity is best, but accelerated ratio, efficiency and cost is not the best. Cannon algorithm and Fox algorithm idea is similar to, but Cannon al- gorithm than the Fox algorithm in data broadcast on cost less, so the overall performance is better. The al- gorithm Systolic based on Pipelining parallel algorithms for matrix multiplication, the better comprehen- sive properties.
作者 陈鹏 樊小超
出处 《新疆师范大学学报(自然科学版)》 2012年第3期5-10,共6页 Journal of Xinjiang Normal University(Natural Sciences Edition)
关键词 矩阵乘并行算法 时间复杂度 性能分析 Matrix multiplication parallel algorithm Time complexity Performance analysis
  • 相关文献

参考文献5

  • 1Dongorra,J. J. ,R. A. Van de Geijn and D. W. Walker. Scalability Issue Affecting the Dsign of Dense Linear Algebra Library [J]. Paralle and Distributed Comuping, 1994,3(22) :513-537.
  • 2Dekel Eliezer, Nassimi David,Sahni Sartaj. Parallel matrix and graph algorithms [J]. SIAMJ. Comput. 1981,4(10) :657-675.
  • 3Alpatov P. , G. Baker, C. Edwards, J. Gunnels, G. Morrow, J. Overfelt, Robert van de Geijn , J. wu. Plapack: Parallel Linear Algebra package [C]. Pcoceedings of the SIAM Parallel Prosseing Conference, 1997.
  • 4Agarwal,R. C. ,S. M. Balle,F. G. Gustavson,M. Joshi,P. Palkar,A 3-Dimensional Approach to Paralle Matrix Multiplication[C] ,IBM J. Res. Develop. 1995,5(22) :1-8.
  • 5Choi J. ,J. J. Dongarra ,D. W. Walker. Level 3 BLAS for distributed memory concurrent computers[J]. CNRS-NSF Workshop on Environments and Tools for Parallel Scientific Computing. 1992(sept. ):7-8.

同被引文献12

引证文献2

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部