期刊文献+

面向多核的并行编程和优化研究 被引量:11

ON PARALLEL PROGRAMMING AND OPTIMISATION FOR MULTI-CORE
下载PDF
导出
摘要 随着多核乃至众核平台的普及,面向多核的并行编程和优化已成为计算机领域研究的热点。然而,绝大多数程序员还依然延续着传统的串行编程习惯,而且目前的主流算法仍以串行为主。因此,如何有效地将串行程序并行化和如何高效地编写多核程序成为多核编程领域亟待解决的问题。对多核编程和优化技术的现状进行全面的研究和分析,在论述如何将串行程序并行化的同时,分析现今主流的一些多核并行编程工具和模型。在此基础上,进一步讨论了在多核编程过程中影响程序性能的因素,并阐述了软硬件领域针对多核编程所做的优化。在对各个研究项目进行分析和评价的基础上,也对面向多核的并行编程和优化技术可能的发展方向进行了展望。 With the popularity of multi-core and even many-core platforms, parallel programming and optimisation for multi-core have be- come the focuses of research in computer science area. However, most of the programmers are still go on the traditional serial programming habits, therefore how to effectively parallelise the serial programs and to efficiently compile the multi-core programs become the issues that need to be urgently resolved. We make the overall studies and analyses on the status quo of multi-core programming and optimisation technolo- gies in the paper. While describing the way to,parallelise the serial programs, we also analyse the tools and models for multi-core parallel pro- gramming which are of the mainstream nowadays. Based on that, we further discuss the factors in multi-core programming process that may af- fect the programs performance, and expatiate on the optimisations made for multi-core programming in both software and hardware area. On the basis of analysing and appraising various research projects, we also present the prospects on possible development direction in regard to parallel programming and optimisation technologies for muhi-core.
出处 《计算机应用与软件》 CSCD 北大核心 2013年第12期198-202,279,共6页 Computer Applications and Software
基金 国家自然科学基金项目(60903015) 国家高技术研究发展计划项目(2009AA012201)
关键词 并行编程 多核 并行工具 并行模型 Parallel programming Multi-core Parallel tool Parallel model
  • 相关文献

参考文献33

  • 1Flynn M. Some computer organizations and their effectiveness [ J ].IEEE Transactions on Computers, 1972,21 ( 9 ) :948 - 960.
  • 2Garland M, Grand S L, Nickolls J, et al. Parallel computing experi- ences with cuda [ J ]. IEEE MICRO, 2008,28 (4) : 13 - 27.
  • 3Kasim H, March V, Zhang R, et al. Survey on Parallel Programming Model[ J]. NPC 2008, LNCS 5245, 2008:266 -275.
  • 4http ://software. intel, com/sites/products/documentationJhpc/tbb/get- ring_started, pdf.
  • 5Frigo M, Leierson C E, Randall K H. The implementation ofthe Cilk- 5 multi-threaded language [ C ]//Proceedings of the1998 ACM SIGP- LAN conference on Programming language design and implementation, 1998:212 - 223.
  • 6Cilk + + : A quick, easy and reliable way to improve threaded perform- ance [ J/OL]. http://software, intel, com/en-us/articles/intel-cilk- plus/.
  • 7Frigo M, Halpem P, Leiserson C E, et al. Reducersand other Cilk ++ hyperobjects[ C]//Proceedings of the twenty-firstannual symposi- um on Parallelism in algorithms and architectures ,2009:79- 90.
  • 8Perez J M, Badia R M, Labarta J. A dependency-aware task-based programming environment for multicore architectures [ C ]//Proceed- ings of the IEEE International Conference on Cluster Computing ( CLUSTER), 2008:142 - 151.
  • 9Thies W, Karczmarek M, Amarasinghe S P. Streamh:A language for streaming applications [ C ]//Proc. of the llth Intl. Conference on Compiler Construction ( CC), 2002 : 179 - 196.
  • 10Buck I, Foley T, Horn D, et al. Hanrahan. Brook for gpus: stream computing on graphics hardware [ C ]//ACM SIGGRAPH 04,2004 : 777 - 786.

同被引文献72

引证文献11

二级引证文献18

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