首页 | 本学科首页   官方微博 | 高级检索  
     

基于GPU的串匹配算法研究
引用本文:张庆丹,戴正华,冯圣中,孙凝晖.基于GPU的串匹配算法研究[J].计算机应用,2006,26(7):1735-1737.
作者姓名:张庆丹  戴正华  冯圣中  孙凝晖
作者单位:1. 中国科学院,计算技术研究所,北京,100080;中国科学院,研究生院,北京,100080
2. 中国科学院,计算技术研究所,北京,100080
基金项目:国家高技术研究发展计划(863计划);中国科学院科研项目
摘    要:BF算法是串匹配算法中最基础的算法,但它是串行算法,不适合图形处理器(Graphic Processing Unit, GPU)的体系结构。结合GPU的特殊体系结构,通过数据存取方式和计算策略的改进,充分利用了GPU的并行处理能力,从而基于GPU实现了BF算法。实验结果表明基于GPU的并行算法能够取得较好的加速比,同时也给出了在现有GPU架构上有效实现通用计算的瓶颈。

关 键 词:图形处理器  通用计算  串匹配  并行
文章编号:1001-9081(2006)07-1735-03
收稿时间:2006-01-12
修稿时间:2006-01-122006-03-10

String matching algorithm research based on graphic processing unit
ZHANG Qing-dan,DAI Zheng-hua,FENG Sheng-zhong,SUN Ning-hui.String matching algorithm research based on graphic processing unit[J].journal of Computer Applications,2006,26(7):1735-1737.
Authors:ZHANG Qing-dan  DAI Zheng-hua  FENG Sheng-zhong  SUN Ning-hui
Affiliation:1. Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100080, China; 2. Graduate School, Chinese Academy of Sciences, Beifing 100080, China
Abstract:BF algorithm is the basic algorithm in string matching field,but it is not parallel,and does not match the system architecture of the Graphic Processing Unit(GPU).Concerning the specific hardware architecture of the GPU,new method of data accessing and computing strategy of the parallel BF algorithm was designed.And by fully using the parallel capaticity of the GPU,the GPU-based algorithm can outperform the CPU-based algorithm.The bottlenecks of GPU architectures were also presented.
Keywords:Graphic Processing Unit(GPU)  general purpose computation  string matching  parallel
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机应用》浏览原始摘要信息
点击此处可从《计算机应用》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号