期刊文献+

基于修正LRU的压缩Cache替换策略 被引量:3

Replacement Policy for Compressed Cache Based on Modified LRU
下载PDF
导出
摘要 以优化压缩cache的替换策略为目标,提出一种优化的基于修正LRU的压缩cache替换策略MLRU-C。MLRU-C策略能利用压缩cache中额外的tag资源,形成影子tag机制来探测并修正LRU替换策略的错误替换决策,从而优化压缩cache替换策略的性能。实验结果表明,与传统LRU替换策略相比,MLRU-C平均能降低L2压缩cache失效率12.3%。 This paper proposes an improved compressed cache replacement policy Modified LRU Replacement Policy for Compressed Cache(MLRU-C) based on modified Least recently Used(LRU) policy for improving compressed cache replacement policy. MLRU-C can use the extra tag in compressed cache to construct shadow ta.g mechanism, which can detect, predict, and correct bad replacement decisions made by LRU policy so as to improve performance of compressed cache replacement policy. Experimental results show MLRU-C can decrease L2 compressed cache miss rate by 12.3% compared with conventional LRU replacement policy.
出处 《计算机工程》 CAS CSCD 北大核心 2008年第18期7-9,16,共4页 Computer Engineering
基金 国家自然科学基金资助项目(90207011)
关键词 影子tag机制 压缩cache 替换策略 shadow tag mechanism compressed cache replacement policy
  • 相关文献

参考文献5

  • 1Alameldeen A R, Wood D A. Frequent Pattem Compression: A Significance-based Compression Scheme for L2 Caches[R]. Computer Sciences Department, University of Wisconsin-Madison, Technical Report: 1500, 2004-04.
  • 2Chen David, Peserico E, Rudolph L. A Dynamically Partitionable Compressed Cache[C]//Proceedings of the Singapore-MIT Alliance Symposium. Singapore: [s. n.], 2003-01.
  • 3Alameldeen A R, Wood D A. Adaptive Cache Compression for High-performance Processors[C]//Proceedings of the 31st Annual International Symposium on Computer Architecture. [S. l.]: IEEE Press, 2004 : 212-223.
  • 4Hallnor E, Reinhardt S. A Unified Compressed Memory Hierarchy[C]//Proc. of the 11th Symposium on High Performance Computer Architecture. [S. l.]: IEEE Press, 2005.
  • 5Belady L A. A Study of Replacement Algorithms for Virtual Storage Computers[J]. IBM Systems Journal, 1996, 5(2): 78-101.

同被引文献17

  • 1付雄,张昱,陈意云.基于复用距离的cache失效率分析[J].小型微型计算机系统,2006,27(9):1777-1781. 被引量:1
  • 2Seongbeom Kim, Dhruba Chandra, Yan Solihin. Fair cache sha- ring and partitioning on a chip multi-processor architecture[ C]. In Proc. of the Intl. Conf. on Parallel Architecture and Compilation Techniques, 2004.
  • 3Soares Livio, Tam David. Reducing the harmful effects of last-lev- el cache polluters with an OS-level, software-only pollute buffer C 1- In: MICRO 41 Proceedings of the 4lst Annual IEEE/ACM International Symposium on Microarchitecture, Washington D C, USA: IEEE Computer Society, 2008: 256-269.
  • 4Erik Berg, Erik Hagersten. StatCache: a probabilistic approach to efficient and accurate data locality analysis [ C ]. In : Performance Analysis of System and Software, 2004 IEEE International Sympo- sium on ISPASS, Washington D C, USA: IEEE Computer Socie- ty, 2004: 20-27.
  • 5Yah Solihin, Fei Guo, Seongbeom Kim. Predicting cache space contention in utility computing servers [ C ]. In: Parallel and Dis- tributed Processing Symposium, 19th IEEE International, Washing- ton D C, USA: IEEE Computer Society, 2005: 8-13.
  • 6Mazen Kharbutli, Keith Irwin, Yan solihin, et al. Using prime numbers for cache indexing to eliminate conflict misses [ C ]. In Proe. of the Intl. Syrup. on High Performance Computer Architec- ture, 2004.
  • 7Jeffrey D Gee, Alan Jay Smith. Analysis of multiprocessor memory reference behavior[ J]. Computer Design: VLSI in Computers and Processors, 1994.
  • 8Radford M Neal. Probabilistic inference using markov cain monte carlo methods[ D]. Department of Computer Science University of Toronto, 1993.
  • 9Christian Beecks, Merih Seran Uysal, Thomas Seidl. Similarity at- fix compression for efficient signature quadratic form distance com- putation[D]. Proceedings of the Third International Conference on Similarity Search and Applications, 2010, 9( 181- 9) :109-114.
  • 10曾浩,叶芃,王厚军,向川云.数字三维示波器波形映射技术研究[J].仪器仪表学报,2009,30(11):2399-2404. 被引量:8

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