期刊文献+

OpenSPARC T1处理器Cache的优化研究

Optimization of the Cache in OpenSPARC T1 processor
下载PDF
导出
摘要 文章以OpenSPARC T1处理器为例,分析了片上多线程结构(chip multi-threading,CMT)处理器由于Cache抖动引发的缓存冲突等问题,通过引入空间锁环机制,减少程序中循环体被替换出Cache的概率,降低Cache冲突,从而提高多线程处理器性能。结果表明,使用空间锁环机制有效降低了缓存延迟和Cache的失效率。 Taking OpenSPARC T1 processor as an example ,the Cache conflicts caused by the Cache thrashing in the chip multi-threading(CM T ) processors are discussed .On this basis ,the circular loc-king mechanism of space is introduced to reduce the probability of being replaced from the Cache and avoid Cache conflicts in the loop body ,so as to improve the performance of multi-threading processor . The result shows that the cache latency and the failure rate of the Cache are reduced by using the cir -cular locking mechanism .
出处 《合肥工业大学学报(自然科学版)》 CAS CSCD 北大核心 2016年第6期786-789,共4页 Journal of Hefei University of Technology:Natural Science
基金 国家自然科学基金资助项目(61106020) 合肥工业大学大学生创新性实验计划资助项目(201510359038)
关键词 OpenSPARC T1 处理器 CACHE 抖动 空间锁环 OpenSPARC T1 processor Cache thrashing circular locking mechanism of space
  • 相关文献

参考文献4

二级参考文献15

  • 1段然,樊晓桠,荆元利,沈戈,梁婕.面向DSP应用的可重构计算(英文)[J].微电子学与计算机,2005,22(4):45-50. 被引量:2
  • 2王熹微,唐昆,崔慧娟.基于DM642的视频编码Cache优化策略[J].微计算机信息,2005,21(09Z):84-86. 被引量:21
  • 3OpenSPARC TI Microarchitecture Specification.
  • 4A. Malik, B. Moyer, D. Cermak, "A Low Power Unified Cache Architecture Providing Power and Performance Flexibility," Int. Symp. on Low Power Electronics and Design, June 2000.
  • 5Austin T, Ernst D,Larson E,et al.Simplescalar Tutorial v4.0.
  • 6Glen Reinman and N.P. Jouppi. “CACTI2.0: An Integrated Cache Timing and Power Model,” 1999. COMPAQ Western Research Lab.
  • 7马鹏云.CMT处理器高速缓存的优化技术[D].长沙:国防科技大学,2007.
  • 8Mutlu O, Stark J, Wilkerson C, et al. Runahead execu- tion: an alternative to very large instruction windows for out - of - order processors[ C]//Proc. the 9th International Symposium on High- Performance Computer Architec- ture. San Jose. USA, 2003 : 129.
  • 9Ganusov I, Burtscher M. Future execution: a hardware prefetchingtechnique for chip multiprocessors [ C]//Proc. The 14th Parallel Architectures and Compilation Tech- niques. St. Louis, USA, 2005:350 - 360.
  • 10Zhou H. Dual - core execution: building a highly scalable single- thread instruction window[ C]//Proc. the 14th Parallel Architectures and Compilation Techniques. St. Louis, USA, 2005,17 (21) : 231 - 242.

共引文献2

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