一种用于异构环境中任务调度的高效算法

High Performance Algorithm for Task Scheduling in Heterogeneous Environment

在异构计算环境中,有效的任务调度对于获得高性能是十分重要的。现在虽然已经有许多异构处理器调度算法,但它们或者不具有良好的效果,或者算法代价太高。提出了一种新的基于表的调度算法APS。APS利用有向无环图来计算任务优先级,并采用基于调度的策略分配任务到不同处理器,以获得任务最少完工时间。将APS和LMT,HEFT,CPOP算法做比较之后得出:在大多数情况下APS算法都能获得更好性能。

Efficient task scheduling is critical for obtaining high performance in a heterogeneous computing environment.Although there are many scheduling algorithms already,they may not have good results,and take high cost.In this paper,a new list scheduling heuristics,named Accurate Priority Scheduling（APS）,is presented.The APS selects task with random directed acyclic graph（DAG） and assigns tasks to processors with the insertion based scheduling policy to minimize the makespan.Compared with LMT,HEFT,CPOP based on,show the APS results good performances in most situations.