期刊文献+

多核系统性能检测及调优策略研究 被引量:1

Research on Performance Detection and Optimization Strategy for Multi-core Microprocessor
下载PDF
导出
摘要 随着电力信息化的发展,各种在线的生产、营销系统的性能已经成为电力系统发展的瓶颈。引入电力系统分析计算从理论上解决了信息系统实时快速分析的难题,但在实际的系统应用中,多核硬件的处理能力并没有完全发挥出来,已渐渐成为整个系统的性能瓶颈。文章结合电力系统软件的特点,采用了基于PMU的多核检测方法,设计并实现了多核环境下电力信息系统性能检测及调优系统软件,分析和检测在线生产、营销系统的瓶颈,找出影响系统性能的主要原因,并生成相应的调优报告,有针对性地提出了系统调整方案。 With the development of electric information technology, the performance of various online production and marketing systems has been a bottleneck in the development of power systems. The introduction of power system analysis and calculation will be helpful to solve the issue of real-time analysis of information systems theoretically. However, the processing power of multi-core hardware does not fulfill the practical application of systems, and has gradually become another performance bottleneck of the whole system. According to the characteristics of power system software, this paper designs and implements a system software for performance detection and optimization of power information systems under the multi-core environment through the PMU-based multicore detection methods. The system can not only analyze and detect performance bottleneck of the on-line production and marketing system, but also determine their root causes and make the corresponding optimization report to tune the system.
出处 《电力信息与通信技术》 2014年第1期42-46,共5页 Electric Power Information and Communication Technology
关键词 多核 性能 微体系事件 性能计数器 multi-core performance micro system event performance monitoring unit
  • 相关文献

参考文献8

  • 1朱孝银..针对多核体系结构性能调试的研究[D].复旦大学,2009:
  • 2闫洁..多核平台上的并行程序性能调优技术与工具[D].中国科学技术大学,2010:
  • 3闫洁,徐恒阳,安虹,刘玉,王耀彬.Pview:一种基于PMU的支持并行程序性能分析的新方法[J].计算机科学,2011,38(2):288-292. 被引量:4
  • 4SPRUNT B. The basics of performance monitoring hardware[J].IEEE Micro,. 被引量:1
  • 5Intel Corporation. VTune[EB/OL].http://www.intel.com/software/product s/vtune/,. 被引量:1
  • 6LEVON J. Oprofile[EB/OL].http://oprofile.sourceforge.net/about/,. 被引量:1
  • 7ERANIAN S. Perfmon2[EB/OL].http://perfmon2.sf.net/,. 被引量:1
  • 8MOLNAR I. Perf[EB/OL].http://kemelnewbies.org/Documents,. 被引量:1

二级参考文献12

  • 1Luk C K,Cohn R,et al.Pin:Building Customized Program Ana-lysis Tools with Dynamic Instrumentation[C] ∥Proc.of the ACM SIGPLAN Conference on Programming Language Design and Implementation(PLDI).Chicago,ILN,US,June 2005:190-200. 被引量:1
  • 2Cantrill B M,Shapiro M W,Leventhal A H.Dynamic Instrumentation of Production System[C] ∥Proc.of USENIX ATEC.Boston,MA,US,June 2004:2-2. 被引量:1
  • 3Srivastava A,Eustace A.Atom:A system for building custo-mized program analysis tools[C] ∥Proc.of PLDI.Orlando,Florida,US,1994:196-205. 被引量:1
  • 4Sprunt B.The basics of performance monitoring hardware[J].IEEE Micro,July 2002:64-71. 被引量:1
  • 5Intel Corp.VTune[EB/OL].http://www.intel.com/ soft-ware/products/ vtune/. 被引量:1
  • 6Levon J.Oprofile[EB/OL].http://oprofile.sourceforge.net/about/. 被引量:1
  • 7Eranian S.Perfmon2[EB/OL].http://perfmon2.sf.net/. 被引量:1
  • 8Molnar I.Perf[EB/DK].Linux Kernel Document,2.6.33. 被引量:1
  • 9Ahn D H,Vetter J S.Scalable analysis techniques for microprocessor performance counter metrics[C] ∥Proc.of Conference on Supercomputing.Baltimore,Maryland,US,Nov.2002. 被引量:1
  • 10Sweeney P F,Hauswirth M,Cahoon B,et al.Using hardwar eperformance monitors to understand the behavior of Java applications[C] ∥Proc.of 3rd Virtual Machine Research and Technology Symposium.San Jose,CA,US,May 2004:5-5. 被引量:1

共引文献3

同被引文献4

引证文献1

二级引证文献4

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