基于多核处理器的多任务并行处理技术研究  

RESEARCH ON MULTI-CORE-BASED MULTITASK PARALLEL PROCESSING TECHNOLOGY

作　　者：汪前进[1] 高勇[1] 李存华[1] 

机构地区：[1]淮海工学院计算机工程学院,江苏连云港222005

出　　处：《计算机应用与软件》2012年第7期141-143,153,共4页Computer Applications and Software

摘　　要：目前,多核处理器已经越来越普及,但是现有算法并不能有效地并行处理数据,造成硬件资源的浪费。为了提高多核处理器的数据处理能力,针对求两个字符串的公共子串算法,提出一种改进方法,在VS.NET环境使用TPL(任务并行库),调度改进的算法处理数据。实验结果表明,改进后的算法可以充分利用多核处理器并行处理数据的特点,提高并行加速比,大大提高数据处理效率。Nowadays the multi-core processor is becoming more and more popular, but the existing algorithm can＇ t parallel process data with efficiency, so that there is a waste of hardware resources. In order to improve the data processing ability of the processor, the authors put forward an improved method for the algorithm to obtain the common substring for two strings. Experiment results show that the improved algorithm can make full use of parallel data processing of the multi-core processor, improve its parallel speedup ratio and greatly increase its data processing efficiency.

关 键 词：多核 多任务 并行算法 任务并行库 VS.NET 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]