期刊文献+

PMESI:一种优化进程私有数据访问的缓存一致性协议 被引量:1

PMESI:optimizing memory performance by selectively deactivating cache coherence for private pages
下载PDF
导出
摘要 并行应用程序中绝大部分的访存是对私有数据的访问,在cache一致性协议上不会产生冲突。传统一致性协议没有根据程序私有数据的访问模式进行针对性设计,存在着很大的优化空间。针对以上的问题,提出了一种支持私有状态的cache一致性协议PMESI,通过动态关闭和激活内存空间的cache一致性目录,优化私有内存空间的访问延迟和功耗。通过时钟精确模拟器的测试,PMESI协议优化了程序中54%的访存,并行程序的执行时间平均缩短了9%。 Parallel program has significant percentage of memory requests that target only private data, which does not need to resolve cache coherence conflicts. Yet traditional coherence protocol does not distinguish between shared and private blocks, which leaves much optimization space. An optimized cache coherence protocol, called PMESI, which dynamically deactivates coherence maintenance for private memory space, was suggested. PMESI achieves two distinguishing features: the reduction of memory access latency and system power consumption. Simulation results on the cycle accurate simulator show that 54% memory references can be efficiently optimized and the program execution time is reduced 9% on average.
出处 《国防科技大学学报》 EI CAS CSCD 北大核心 2013年第1期97-102,共6页 Journal of National University of Defense Technology
基金 国家自然科学基金资助项目(60803040 61003301) 国家863计划项目(2012AA01A301)
关键词 PMESI协议 私有内存空间 目录协议 操作系统 PMESI protocol private memory space directory coherence protocol operating system
  • 相关文献

参考文献13

  • 1Conway P, Kalyanasundharam N, Donley G, et al. Cache hierarchy and memory subsystem of the AMD opteron processor [J]. IEEE Micro, 2010,30(2) :16 -29.
  • 2Bias C, Alberto R, Marfa E G, et al. Increasing the effectiveness of directory caches by deactivating coherence forprivate memory blocks [ C ]//38th Int' 1 Symp. on Computer Architecture (ISCA) ,2011.
  • 3Shah M, Barreh J, Brooks J, et al. UltraSPARC T2 : A highly- threaded, power-efficient SPARC SoC [ C ]//IEEE Asian Solid- State Circuits Conference, 2007.
  • 4Hardavellas N, Ferdman M, Falsafi B, et al Reactive NUCA: Near-optimal block placement and replication in distributed caches [ C ]//36th Int Symp. on Computer Architecture ( ISCA ), 2009.
  • 5Kim D, Ahn J, Kim J, et al. Subspace snooping: Fihering snoops with operating system support[ C ]//19th Int Conference Parallel Architectures and Compilation Techniques (PACT) , 2010.
  • 6Magnusson P S, Christensson M,Eskilson J, et al. Simics : A full system simulation platform[ J]. IEEE Computer, 2002,35 (2) :50 -58.
  • 7Martin M M, Sorin D J, Beckmann B M, et al. Muhifacet' s general execution-driven muhipmcessor simulator (GEMS) toolset[J]. Computer Architecture News, 2005,33(4) :92 -99.
  • 8Bienia C, Kumar S, Singh J P,et al. The PARSEC benchmark suite: characterization and architectural implications [ C ]// 17th Int Conference on Parallel Architectures and Compilation Techniques (PACT), 2008.
  • 9Woo S C, Ohara M, Torrie E, et al. The SPLASH-2 programs: Characterization and methodological considerations[C]//22nd Int Symp. on Computer Architecture (ISCA), 1995.
  • 10Ekman M, Dahlgren F, Stenstrfim P. TLB and snoop energy- reduction using virtual caches[ C]//Int Syrup. on Low Power Electronics and Design (ISLPED) , 2002.

同被引文献6

引证文献1

二级引证文献3

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