高级检索

    群机系统上单并发任务簇的近优分配算法

    A SUB\|OPTIMAL ALGORITHM ON ALLOCATING A SINGLE TASK CLUSTER ON NOWs

    • 摘要: 缩短程序的执行时间是并行处理的首要目标,有效的任务分配算法是实现这一目标的关键,对群机系统来说更是如此.文中针对并行语言中常用的并行范式——单并发任务簇提出了近优分配算法 O P T A,并在群机系统上做了与 M H 算法的比照实验,结果表明较 M H 算法缩短程序执行时间 10% 左右

       

      Abstract: Reducing the execution time of program is a main goal of parallel processing ,while an efficient task allocation algorithm is a crucial step,especially on NOWs. A sub\|optimal algorithm OPTA on allocating a single task cluster to physical processors is presented in the paper here, and this task cluster is a common parallel paradigm in most parallel programming languages.An experiment of comparing with algorithm MH is made, and the results show that algorithm OPTA cuts 10% cost.

       

    /

    返回文章
    返回