期刊文献+

基于预测缓存的低功耗TLB快速访问机制 被引量:2

Fast and low power TLB access mechanism with prediction buffer
下载PDF
导出
摘要 基于存储器访问局部性原理,提出了一种基于预测缓存的低功耗转换旁置缓冲器(TLB)快速访问机制。该机制采用单端口静态随机存储器(SRAM)代替传统的内容寻址存储器(CAM)结构,通过匹配搜索实现全相连TLB的快速访问,在两级TLB之间设计可配置的访问预测缓存,用于动态预测第二级TLB访问顺序,减少第二级TLB搜索匹配的延时,并有效降低第二级TLB访问功耗。采用该机制明显降低了TLB的缺失代价,当第一级TLB缺失时访问第二级TLB的平均访问延时接近1个时钟周期,约为原有平均访问延时的20%,增加的面积开销仅为原内存管理单元的1.81%左右,具有低成本、低功耗的特征。 This paper proposed a fast and low power TLB access mechanism with prediction buffer based on memory access locality principle,and designed a two-level TLB structure implemented by SARM instead of CAM to achieve fast access of the full associated TLB.Between the two levels of the introduced TLB,an independent and hardware configurable prediction buffer was designed to dynamically predict the access sequences of the second level TLB,which could reduce its access penalty when the first level TLB missed and significantly reduce the dynamic power consumption with little control logic.Experiment shows that compared with the traditional two-level TLB structure,the average access cycles of the second level TLB are about 20% of the traditional one,with only 1.81% area increment,which support low power and low cost embedded application.
出处 《计算机应用研究》 CSCD 北大核心 2011年第8期2964-2966,2996,共4页 Application Research of Computers
基金 国家"863"高科技研究发展计划资助项目(2004AA1Z1020)
关键词 内存管理单元 两级转换旁置缓冲器 内容寻址存储器 静态随机存储器 预测缓存 快速访问 低功耗 MMU two-level TLB CAM SRAM prediction buffer fast access low-power
  • 相关文献

参考文献14

  • 1EKMAN M, DAHLGREN F, STENSTROM P. TLB and snoop ener- gy-reduction using virtual caches in low-power chip-multiprocessors [ C ]//Proe of the 2002 International Symposium on Low Power Electronics and Design. New York: ACM Press,2002:245-246.
  • 2SRIBALAN T, SANTHANM. Strong, ARM SAII0, a 160Mhz 32h 0. 5w CMOS ARM processor[ C]// Hot Chips 8: A Symposium on High-Performance Chips. 1996 : 119-130.
  • 3LEE J H, WEEMS C, KIMS D. Selective block buffering TLB system for embedded proccssom[ J]. IEEE Proceedings Computers and Digital Techniques ,2005,152(4) :507-516.
  • 4LEE J H, LEE J S, JEONG S W, etal. A banked-promotion TLB for high performance and low power[ C ]//Proc of International Conference on Computer Design. 2001:118-123.
  • 5MANNE S, KLAUSER A, GRUNWALD D, etal. Low power TLB design for high performance microprocessors[ R]. [ S. 1. ] : University of Colorado, 1997.
  • 6TALLURI M, HILL M D. Surpassing the TLB performance of superpagees with less operating system support [ C]//Proc of International Conference on Architectural Support for Programming Languages and Opolrating Systems -ASPLOS. San Jose, CA : ACM Press, 1994 : 171.
  • 7LEE J-H, JEONG S-W, KIM S-D, et al. A banked-promotion translation'lookaside buffer system [ J ]. Journal of Systems Architecture ,2002,47 ( 14-15 ) : 1065-1078.
  • 8JOUPPI N P. Imlamving direct-mapped cache performance by the addition of a small fully associative cache and prefetch buffers [ C ]// Proc of the 17th ISCA. New York: ACM Press,1990:364-373.
  • 9CHOIJ H, LEE J H, PARK G H, et al. An advanced filtering TLB for low power coresumption[ C ]//Proe of the 14th Symposium on Computer Architecture and High Performance Computing. Washington DC: IEEE Computer Society,2002:93-99.
  • 10KIN J, GUPTA M, MANGiONE-SM1TH W H. The filter cache: anenergy efficient memory structure[ C ]//Proc of the 30th Annual IEEE International Symposium on Micrnarchitecture. Washington DCi IEEE Computer Society, 1997 : 184-193.

二级参考文献8

  • 1张宇弘,王界兵,严晓浪,汪乐宇.标志预访问和组选择历史相结合的低功耗指令cache[J].电子学报,2004,32(8):1286-1289. 被引量:6
  • 2潘国振,王界兵,严晓浪.高性能嵌入式CPU特殊指令单元的设计与实现[J].浙江大学学报(工学版),2005,39(2):211-215. 被引量:3
  • 3AUSTIN T M,SOHI G S.High-bandwidth address translation for multiple-issue processors[C]∥Proceedings of the 23rd Annual International Symposium on ComputerArchitecture.Philadelphia,US:[s.n.],1996:158-167.
  • 4Tensilica Inc.Xtensa microprocessor overview handbook:a summary of the Xtensa microprocessor data book[EB/OL].[2006-01-15].http://www.tensilica.com/products/white_papers.htm
  • 5MANNE S,KLAUSTER A,GRUNWALD D,et al.Low power TLB design for high performance microprocessors[R].Colorado,US:University of Colorado,1997.
  • 6GUTHAUS M,RINGENBERG J,ERNST D,et al.MiBench:a free commercially representative embedded benchmark suite[C]∥IEEE 4th Annual Workshop on Workload Characterization.Austin,US:IEEE,2001:3-14.
  • 7HENNESSY J L,PATTERSON D A.Computer architecture:a quantitative approach[M].3rd ed.San Francisco:Morgan Kaufmann,2002.
  • 8JACOB B L,MUDGE T N.A look at several memory management units,TLB refill mechanisms and page table organizations[C]∥ ACM the 8th International Conference on Architectural Support for Programming Languages and Operating Systems.San Jose,US:ACM,1998:295-306.

共引文献3

同被引文献14

  • 1崔更申,黄廷辉,王力.一种通用的嵌入式操作系统内存保护设计[J].计算机工程,2005,31(24):96-98. 被引量:2
  • 2刘坤杰,游海亮,严晓浪,葛海通.面向嵌入式应用的内存管理单元设计[J].浙江大学学报(工学版),2007,41(7):1078-1082. 被引量:4
  • 3BARR T W, COX A L, RIXNER S. SpecTLB: a mechanism for speculative address translation [C]//2011 38th Annual Inter- national Symposium on Computer Architecture (ISCA). [S.l.]: IEEE, 2011: 307-317.
  • 4SAULSBURY A, DAHLGREN F, STENSTROM P. Recency- based TLB preloading [J]. ACM SIGARCH Computer Architec- ture News, 2000, 28(2): 117-127.
  • 5CHOI J H, LEE J H, PARK G H, et al. An advanced filtering TLB for low power consumption[ C ]//Proc of the 4th Symposium on Com-puter Architecture and High Performance Computing. [ S. 1. ] : IEEE Press, 2002 : 93- 99.
  • 6LEE J H, JEONG S W, KIM S D, et al. A banked-promotion trans- lation lookaside buffer system [ J ]. Journal of Systems Architec- ture,2002,47(14) :1065-1078.
  • 7CHANG Y J. An ultra low-power TLB design [ C ] //Proc of Confe- rence on Design, Automation and Test in Europe. 2006 : 1122-1127.
  • 8CHANG Y J, LAN M F. Two new techniques integrated for energy- efficient TLB design[J]. IEEE Trans on Very Large Scale Inte- gration ( VLSI ) Systems,2007,15 ( 1 ) : 13- 23.
  • 9LEE J H, WEEMS C, KIM S D. Selective block buffering TLB sys- tem for embedded processors [ J]. IEEE Proceedings Computers and Digital Techniques, lEE Proc,2005,152(4) :507-516.
  • 10HENNESSY J L, PATTERSON D A. Computer architecture: a quantitative approach [ M ]. 4th ed. San Francisco : Morgan Kaufmann Publishers ,2007 : 188-247.

引证文献2

二级引证文献1

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