全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...
电子学报  2012 

一种混合型值关联间接跳转预测机制

DOI: 10.3969/j.issn.0372-2112.2012.11.024, PP. 2298-2302

Keywords: 转移预测,间接跳转,值关联,混合型预测器

Full-Text   Cite this paper   Add to My Lib

Abstract:

准确的间接跳转预测对现代处理器的性能和能耗有效性都具有重要意义.本文提出了一种混合型值关联间接跳转预测机制,通过混合使用多种关联信息以降低间接跳转误预测率.该机制一方面依赖于编译器根据高层次数据流信息识别间接跳转指令所对应的初始关联数据值.另一方面,该机制针对间接跳转预测的不同场景分别设计了两类关联信息:单一数据值和值历史,并实现了一种低开销的硬件结构,该硬件结构在运行时刻根据不同应用场景动态选择最佳关联信息引导间接跳转预测.实验结果表明,相对于传统的BTB预测器和最新的VBBI预测器,本文机制能够有效降低误预测率,提高程序性能并降低系统能耗.

References

[1]  M X Tan,et al.Energy-efficient branch prediction with compiler-guided history stack .Int''l Conference on Design,Automation and Test in Europe .2012.449-454.
[2]  袁楠,范东睿.高性能代价比的两层关联间接转移预测器设计[J].计算机学报,2008,31(11):1898-1906. Yuan N,Fan D R.Design of cost-effective 2-level correlation indirect branch prediction[J].Chinese Journal of Computers,2008,31(11):1898-1906.(in Chinese)
[3]  T H Heil,et al.Improving branch predictors by correlating on data values .Int''l symposium on Microarchitecture .IEEE Computer Society,1999.28-37.
[4]  J A Joao,et al.Improving the performance of object-oriented languages with dynamic predication of indirect jumps .Int''l conference on Architectural Support for Programming Languages and Operating Systems .New York:ACM,2008.80-90.
[5]  D Brooks,et al.Wattch:a framework for architectural-level power analysis and optimizations .Int''l Symposium on Computer Architecture .IEEE,2000.83-94.
[6]  E Perelman,et al.Using SimPoint for accurate and efficient simulation .Int''l conference on Measurement and Modeling of Computer Systems .New York:ACM,2003.318-319.
[7]  M U Farooq,et al.Value based BTB indexing for indirect jump prediction .Int''l symposium on High-Performance Computer Architecture .New York:ACM,2010.1-11.
[8]  朱德新,程旭,慎辉.UNICORE体系结构中动态转移预测机制的研究与设计[J].电子学报,2004,32(8):1351-1355. Zhu D X,Cheng X,Shen H.Dynamic branch prediction research and design for UNICORE architecture [J].Acta Electronica Sinic,2004,32(8):1351-1355.(in Chinese)
[9]  H Kim,et al.VPC prediction:reducing the cost of indirect branches via hardware-based dynamic devirtualization .Int''l Symposium on Computer Architecture .IEEE Computer Society,2007.424-435.
[10]  D R Kaeli and P G Emma.Branch history table prediction of moving target branches due to subroutine returns .Int''l Symposium on Computer Architecture .IEEE Computer Society,1991.34-42.
[11]  J K F Lee and A J Smith.Branch prediction strategies and branch target buffer design[J].IEEE Computer,1984,17(1):6-22.
[12]  Y Chang,et al.Target prediction for indirect jumps .Int''l Symposium on Computer Architecture .1997.274-283.
[13]  S Gochman,et al.The Intel Pentium M processor:microarchitecture and performance[J].Intel Technology Journal,2003,7(2):21-36.
[14]  K Driesen and U H?lzle.The cascaded predictor:economical and adaptive branch target prediction .Int''l symposium on Microarchitecture .IEEE Computer Society,1998.249-258.
[15]  A Seznec and P Michaud,A case for (partially) TAgged GEometric history length predictors[J].Journal of Instruction-Level Parallelism,2006.
[16]  L Chen,et al.Dynamic data dependence tracking and its application to branch prediction .Int''l symposium on High-Performance Computer Architecture .New York:ACM,2003.65-76.
[17]  A Roth,et al.Improving virtual function call target prediction via dependence-based pre-computation .Int''l Conference on Supercomputing .New York:ACM,1999.356-364.
[18]  T Austin,et al.,SimpleScalar:an infrastructure for computer system modeling[J].IEEE Computer,2002.35(2):59-67.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133