一种静态LoC关键性预测器设计

doi:10.3969/j.issn.1000-3428.2012.07.083

计算机工程 ›› 2012, Vol. 38 ›› Issue (7): 253-256. doi: 10.3969/j.issn.1000-3428.2012.07.083

一种静态LoC关键性预测器设计

李清波¹，苟鹏飞¹，孙骏²，杨兵¹，王进祥¹

(1. 哈尔滨工业大学微电子中心，哈尔滨 150001；2. 上海航天控制工程研究所，上海 200072)

收稿日期:2011-08-09 出版日期:2012-04-05 发布日期:2012-04-05
作者简介:李清波(1985－)，男，硕士研究生，主研方向：微处理器体系结构，程序控制指令；苟鹏飞，博士研究生；孙骏，硕士研究生；杨兵，博士后；王进祥，教授、博士生导师

Design of Static LoC Key Predictor

LI Qing-bo ¹, GOU Peng-fei ¹, SUN Jun ², YANG Bing ¹, WANG Jin-xiang ¹

(1. Microelectronics Center, Harbin Institute of Technology, Harbin 150001, China; 2. Shanghai Institute of Control Engineering for Aerospace, Shanghai 200072, China)

Received:2011-08-09 Online:2012-04-05 Published:2012-04-05

摘要/Abstract

摘要： 针对不同分簇超标量处理器结构下SPEC2000程序中指令关键可能性(LoC)的特性，提出一种静态LoC关键性预测器的设计方法。对指令LoC进行研究，根据其结构无关性和动态不变性，设计预测器。仿真结果表明，在对1×8分簇超标量处理器使用该设计时，程序的每周期指令数平均提升5.3%，性能优于动态LoC预测器。

关键词: 超标量处理器, 结构无关性, 动态不变性, 静态预测, 指令调度

Abstract: Aiming at instruction Likelihood of Criticality(LoC) of SPEC2000 programs under different clustered superscalar processor architectures, this paper proposes a design of static LoC key predictor. It discusses the instruction LoC, and finds that it has some characteristics, such as architecture-independent and dynamic-invariable. This paper uses this characteristics above to design the predictor. Simulation results show that when 1×8 cluster super-scalar processor uses the design, the program Instruction per Clock(IPC) average increases by 5.3%, and the perfor- mance is better than the dynamic LoC predictors.

Key words: superscalar processor, structure independence, dynamic invariability, static prediction, instruction schedule

中图分类号:

TP311.5

李清波, 苟鹏飞, 孙骏, 杨兵, 王进祥. 一种静态LoC关键性预测器设计[J]. 计算机工程, 2012, 38(7): 253-256.

LI Qing-Bei, GOU Feng-Fei, SUN Jun, YANG Bing, WANG Jin-Xiang. Design of Static LoC Key Predictor[J]. Computer Engineering, 2012, 38(7): 253-256.

http://www.ecice06.com/CN/Y2012/V38/I7/253

参考文献

[1] Fields B A, Rubin S R B. Focusing Processor Policies via Critical-path Prediction[C]//Proc. of the 28th Annual International Symposium on Computer Architecture. Gothenburg, Sweden: [s. n.], 2001.
[2] Tune E, Dongning L, Tullsen D M, et al. Dynamic Prediction of Critical Path Instructions[C]//Proc. of the 7th International Symposium on High-performance Computer Architecture. [S. l.]: IEEE Press, 2001.
[3] 李清波. 处理器中指令关键性的研究与应用[D]. 哈尔滨: 哈尔滨工业大学, 2011.
[4] Salverda P, Tucker C, Zilles C. Accurate Critical Path Prediction via Random Trace Construction[C]//Proc. of the 6th Annual IEEE/ACM International Symposium on Code Generation and Optimization. Boston, USA: [s. n.], 2008.
[5] Salverda P, Zilles C. A Criticality Analysis of Clustering in Superscalar Processors[C]//Proc. of the 38th Annual IEEE/ACM International Symposium on Microarchitecture. Barcelona, Spain: [s. n.], 2005.
[6] Parcerisa J M. Design of Clustered Superscalar Microarchi- tectures[D]. Barcelona, Spain: Politècnicade Catalunya, 2004.
[7] Sherwood T, Perelman E, Hamerly G, et al. Discovering and Exploiting Program Phases[J]. IEEE Computer Society, 2003, 23(6): 84-93.
[8] Srinivasan S T, Lebeck A R. Load Latency Tolerance in Dynamically Scheduled Processors[C]//Proc. of the 31st Annual IEEE/ACM International Symposium on Microarchitecture. Dallas, USA: [s. n.], 1998.
[9] Grunwald D, Klauser A, Manne S, et al. Confidence Estimation for Speculation Control[C]//Proc. of the 25th Annual International Symposium on Computer Architecture. Barcelona, Spain: [s. n.], 1998.
[10] Robatmili B, Govindan S, Burger D, et al. Exploiting Criticality to Reduce Bottlenecks in Distributed Uniprocessors[C]//Proc. of the 17th International Symposium on High Performance Computer Architecture. San Antonio, USA: [s. n.], 2011.
[11] Binkert N L, Dreslinski R G, Hsu L R, et al. The M5 Simulator: Modeling Networked Systems[J]. IEEE Computer Society, 2006, 26(4): 52-60.
[12] 杨兵. 分簇超标量处理器关键技术研究[D]. 哈尔滨: 哈尔滨工业大学, 2009.
[13] Perelman E, Hamerly G, Calder B. Picking Statistically Valid and Early Simulation Points[C]//Proc. of the 12th International Conference on Parallel Architectures and Compilation Techniques. New Orleans, USA: [s. n.], 2003.
[14] Dean J, Hicks J E, Waldspurger C A, et al. ProfileMe: Hardware Support for Instruction-level Profiling on Out-of-order Processors[C]//Proc. of the 13th Annual IEEE/ACM International Symposium on Microarchitecture. [S. l.]: IEEE Press, 1997.
[15] Riley N, Zilles C. Probabilistic Counter Updates for Predictor Hysteresis and Stratification[C]//Proc. of the 12th International Symposium on High-performance Computer Architecture. [S. l.]: IEEE Press, 2006.

选择文件类型/文献管理软件名称

选择包含的内容

一种静态LoC关键性预测器设计

Design of Static LoC Key Predictor

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

编辑推荐

Metrics

本文评价

[1]	李昭, 刘有耀, 焦继业, 潘树朋. 超标量处理器乱序提交机制的研究与设计[J]. 计算机工程, 2021, 47(4): 180-186.
[2]	曹代,郭绍忠,张辛. 基于申威26010处理器的扩展函数库实现与优化[J]. 计算机工程, 2017, 43(1): 61-66,71.
[3]	曹晓，李莹. 基于反馈的JCVM指令预调度方案[J]. 计算机工程, 2014, 40(1): 78-82.
[4]	穆雅莉, 杨兵, 喻明艳. 基于区间模型的一级指令Cache缺失损失分析[J]. 计算机工程, 2012, 38(7): 273-275,278.
[5]	王正华, 郭炜, 魏继增. 基于最小延时启发式搜索的TTA代码优化[J]. 计算机工程, 2010, 36(10): 282-284.
[6]	张垚. 保证Java精确异常的指令调度技术[J]. 计算机工程, 2009, 35(8): 122-124.
[7]	胡维;祝永新;姜雷. 基于整数线性规划的TTA代码优化[J]. 计算机工程, 2008, 34(21): 219-221,.
[8]	史晓华;刘超;金茂忠;郭鹏. 即时编译器中的轻量级指令调度算法[J]. 计算机工程, 2007, 33(15): 3-6.

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

一种静态LoC关键性预测器设计

Design of Static LoC Key Predictor

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

编辑推荐

Metrics

本文评价