期刊文献+

片上多处理器共享Cache的访存时间最优划分方法

Optimal memory time Cache partitioning in chip-multiprocessors
下载PDF
导出
摘要 提出的访存时间最优Cache划分(OMTP,optimal memory time Cache partitioning)方法通过特征获取部件来获取不同应用程序的平均失效开销和Cache命中的路分布情况,以此作为划分依据来给竞争程序分配合适的Cache空间,达到优化程序整体执行性能的目的。实验结果表明,OMTP方法相比基于利用率的Cache划分(UCP)方法吞吐率平均提高3.1%,加权加速比平均提高1.3%,整体性能更优。 Optimal memory time Cache partitioning(OMTP) was proposed.The OMTP can get the average access in-validation overheads of different application and Cache line distributation about Cache hit through the characteristic ob-tain unit.According to which the OMTP can allocate proper Cache capacity to competitive process,so the overall per-formance of the program can achieve optimizied.The experiment results showed that OMTP had higher miss rate com-pared with utility-based Cache partitioning(UCP),but it had better IPC throughput and weighted speedup,OMTP im-proved throughput by average 3.1% and weighted speedup by average 1.3% over UCP.
作者 李浩 谢伦国
出处 《通信学报》 EI CSCD 北大核心 2012年第4期136-142,共7页 Journal on Communications
关键词 片上多处理器 共享CACHE CACHE划分 访存时间最优划分 chip-multiprocessor shared Cache Cache partitioning optimal memory time Cache partitioning
  • 相关文献

参考文献12

  • 1STONE H S. Optimal partitioning of Cache memory[J].IEEE Transaction on Computer,1992,(09):1054-1068.doi:10.1109/12.165388.
  • 2SUH G F,RUDOLPH L,DEVADAS S. Dynamic partitioning of shared Cache memory[J].Journal of Supercomputing,2004,(01).
  • 3QURFSHI M K,PATT Y N. Utility-based Cache partitioning:a low-overhead,high-performance,runtime mechanism to partition shared Caches[A].2006.423-432.
  • 4QURESHI M K,LYNCH D N,MUTLU O. A case for MLP-aware Cache replacement[A].2006.167-178.
  • 5KROFT D. Lockup-free instruction fetch/prefetch Cache organization[A].1981.81-88.
  • 6MUTLU O,STARK J,WILKERSON C. Runahead execution:an alternative to very large instruction windows for out-of-order processors[A].2003.129-140.
  • 7CHOU Y,FAHS B,ABRAHAM S G. Microarchitecture optimizations for exploiting memory-level parallelism[A].2004.76-89.
  • 8HOU X,CHEN W G,ZHENG W M. Cache sharing management for performance fairness in chip nultiprcessors[A].Raleigh,North Carolina,USA,2009.384-393.
  • 9RAJKUMAR R. A resource allocation model for QoS management[A].1997.298-307.
  • 10MARTIN M M K,SORIN D J,BECKMANN B M. Multifacets general execution-driven multiprocessor simulator (gems) toolset[J].SIGARCH Comput Archit News,2005,(04):92-99.

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