计算机科学 ›› 2015, Vol. 42 ›› Issue (11): 94-95.doi: 10.11896/j.issn.1002-137X.2015.11.019

• 2014年全国高性能计算机学术年会 • 上一篇    下一篇

基于MIC的GFSR(521,2)并行化设计与实现

顾小璐,周津羽,华 诚,刘 逍,周晓辉   

  1. 西安邮电大学计算机学院 西安710121,陕西省高性能计算研究中心 西安710121,陕西省高性能计算研究中心 西安710121,西安邮电大学计算机学院 西安710121;陕西省高性能计算研究中心 西安710121,西安邮电大学计算机学院 西安710121;陕西省高性能计算研究中心 西安710121
  • 出版日期:2018-11-14 发布日期:2018-11-14
  • 基金资助:
    本文受陕西省自然科学基础研究计划(2013JM8028),高效能服务器和存储技术国家重点实验室开放课题资助

Implementation and Performance Analysis of GFSR(521,2) Parallelization Based on MIC

GU Xiao-lu, ZHOU Jin-yu, HUA Cheng, LIU Xiao and ZHOU Xiao-hui   

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

摘要: GFSR是一种反馈移位随机数发生器。在研究GFSR(521,2)串行算法的基础上,利用Strided skip ahead方法对其进行并行化。实验结果显示,GFSR(521,2)并行化后TestU01的测试结果与串行的相同。相对于CPU单线程,MIC平台下的最优加速比为7.58。

关键词: 随机数发生器,GFSR,并行化,MIC

Abstract: The GFSR is a kind of feedback shift random number generator.Based on the study of GFSR (521,2) serial algorithm, we used Strided skip ahead method to realize the parallelization.Experimental results show that the paralleli-zed GFSR (521,2) generator’s TestU01 test results are the same as serial algorithm’s.And the best speedup based on the MIC platform reaches 7.58 relative to single-thread with the CPU.

Key words: Random number generator,GFSR,Parallelization,MIC

[1] L’Ecuyer P,Tezuka S.Structural properties for two classes of combined random number generators[J].Mathematics of Computation,1991,57(196):742-743
[2] L’Ecuyer P,Blouin F,Couture R.A search for good multiple recursive random number generators[J].ACM Transactions on Modeling and Computer Simulation,1993,3(2):87-98
[3] Bradley T,du Toit,Giles J,et al.Parallelization techniques for random number generators[M]∥GPU Computing Gems Eme-rald Edition.2011:231-246
[4] Makino J,Miyamura O.Generation of Shift Register RandomNumbers on Vector Processor[J].Computer Physics Communication,1991,64(23):363-368
[5] Makino J,Takaishi T,Miyamura O.Generation of Shift Register Random Numbers on Distributed Memory Multiprocessors[J].Computer Physics Communication,1992,70(3):495-500
[6] Wei Gong-yi,Yang Zi-qiang.Some algorithms of parallel random number generators[J].Journal of Numerical Methods and Computer Applications,2001(4):311-320
[7] Lewis T G,Payne W H.Generalized Feedback Shift RegisterPseudorandom Number Algorithm[J].Journal of the ACM,1973,20(3):457-460
[8] Ripley B D.Thoughts on pseudorandom number generators[J].Journal of Computational and Applied Mathematics,1990,31(1):156-157
[9] Wang En-dong,Zhang Qing,Shen Bo,et al.High-Performance Computing on the Intel Xeon Phi-How to Fully Exploit MIC Architectures[M].China Water Power Press,2012
[10] L’Ecuyer P,Simard R.TestU01:A C Library for EmpiricalTesting of Random Number Generators[J].ACM Transactions on Mathematical Software,2007,33(4):22

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!