期刊文献+

片上多处理器中时空结合的数据预取

Spatio-temporal Data Prefetching for Chip Multiprocessors
下载PDF
导出
摘要 片上多处理器中不同核的缺失地址序列之间通常存在一定的空间和时间相关性,为了充分利用该性质,本文提出时空结合的数据预取.该机制首先寻找核内缺失地址序列的相关性,在核内探索不到的情况下再寻找核间的相关性,因此可利用其它核的访存行为来预测本核可能即将发生的访存行为.实验结果表明,本文提出的数据预取机制可使测试程序的平均性能提高12.6%,与扩展应用在多核上的C/DC策略相比较,性能提高了3.8%. The miss addresses of multiple cores often have some spatial correlation and temporal correlation in chip multiprocessors.In order to make full use of the properties,this paper proposes Spatio-Temporal data prefetching which exploits inter-core correlation when it don′t find intra-core correlation in a core,so it can predict future memory access through other cores.The experimental results indicate that the proposed mechanism can get an average of 12.6% speedup when compared with no prefetching.Meanwhile,compared with C/DC,it improves performance by 3.8%.
出处 《小型微型计算机系统》 CSCD 北大核心 2012年第8期1856-1861,共6页 Journal of Chinese Computer Systems
基金 国家自然科学基金项目(61070001)资助
关键词 片上多处理器 空间相关 时间相关 数据预取 chip multiprocessors spatial correlation temporal correlation data prefetching
  • 相关文献

参考文献10

  • 1Renau J, Fraguela B, Tuck J, et al. SESC simulator [EB/OL]. http:llsesc. sourceforge. net,2005.
  • 2Woo S C, Ohara M, Torrie E,et al. The SPLASH-2 programs: characterization and methodological considerations [C]. Computer Architecture, ISCA 1995, Santa Margherita Ligure, Italy, ACM, 1995.
  • 3Joseph D, Grunwald D. Prefetching using markov predictors [C]. ISCA 1997 ,Denver, USA, ACM , 1997 :252-263.
  • 4Kandiraju G B, Sivasubramaniam A. Going the distance for TLB prefetching , an application-driven study [C]. ISCA 2002, Anchorage, Alaska, IEEE Computer Society, May,2002:195-206.
  • 5Nesbit K J, Smith J E. Data cache prefetching using a global history buffer [A]. 10th International Conference on High-Performance Computer Architecture [C]. Madrid, Spain, IEEE Computer Society, 2004 :96-106.
  • 6Nesbit K J , Dhodapkar AS, Smith J E. AC/DC: an adaptive data cache prefetcher [A). 13th International Conference on Parallel Architectures and Compilation Techniques [C). Antibes Juan-lesPins, IEEE Computer Society, 2004 :135-145.
  • 7Somogyi S, Wenisch T F, Ailamaki A, et al. Spatial memory streaming [C]. ISCA 2006 , Boston, MA, USA, IEEE Computer Society, 2006: 252-263.
  • 8Wenisch T F, Somogyi S, Hardavellas N , et al. Temporal streaming of shared memory [A]. The 32st International Symposium on Computer Architecture [C]. Madison, Wisconsin, USA, IEEE Computer Society, 2005 :222-233.
  • 9Palacharla S, Kessler R E. Evaluating stream buffers as a secondary cache replacement [A]. Proceedings of 21 st International Symposium on Computer Architecture [C]. Chicago, IL, USA, IEEE Computer Society Press ,1994.
  • 10Diaz P, Cintra M. Stream chaining: exploiting multiple levels of correlation in data prefetching [C]. In Proceedings of the 36th International Symposium on Computer Architecture, Austin, Texas, USA, ACM,2009 :81-92.

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