本文已被：浏览 2091次 下载 4638次
Received:June 20, 2011 Revised:July 23, 2011
Received:June 20, 2011 Revised:July 23, 2011
中文摘要: 多模式匹配是计算机科学中最基本的问题,其应用在许多领域,在一些情形下也是比较耗时的。GPU拥有比CPU 更强的并行计算能力,随着CUDA 架构的推出,GPU 用于通用计算领域的并行编程工作变得更加轻松。实现了基于CUDA 架构的Wu-Manber 多模式匹配算法,实验结果表明,相比传统串行算法而言,本文的实现获得了10 倍以上的加速。
Abstract:Multi-pattern matching is a basic problem in computer science and used in many fields, in some cases, also the most time-consuming. GPU has more parallel computing capabilities than the CPU. With the introduction of CUDA, GPU computing for general purpose parallel programming becomes easier. This paper proposes Wu-Manber multi-pattern matching algorithm based on the CUDA, and evaluating the implementations we have achieved speedups up to 10 faster than the sequential implementations.
keywords: multi-pattern matching GPU CUDA Wu-Manber
文章编号： 中图分类号： 文献标志码：
基金项目:
引用文本：
马计,王国平,杨明.基于CUDA 的Wu-Manber 多模式匹配算法.计算机系统应用,2012,21(3):51-54,175
MA Ji,WANG Guo-Ping,YANG Ming.Wu-Manber Multi-pattern Matching Algorithm Based on CUDA.COMPUTER SYSTEMS APPLICATIONS,2012,21(3):51-54,175
马计,王国平,杨明.基于CUDA 的Wu-Manber 多模式匹配算法.计算机系统应用,2012,21(3):51-54,175
MA Ji,WANG Guo-Ping,YANG Ming.Wu-Manber Multi-pattern Matching Algorithm Based on CUDA.COMPUTER SYSTEMS APPLICATIONS,2012,21(3):51-54,175
