摘要
现有循环并行识别方法用于众核处理器时存在一定不足,当选择的循环并行维迭代数较少时可能导致严重的负载不均衡。针对这一问题,提出了一种面向众核处理器的多维并行识别方法。在现有并行识别方法无法做到较好的负载均衡时,选择嵌套循环的多个维进行并行,将多个并行维的迭代空间合并后再作任务划分,减少负载不均衡对程序并行效率的影响。此方法在已开发的自动并行化系统中进行了实现,实际应用过程中能够很好地提升一些应用程序在众核处理器上并行执行的效率。
There were some shortcomings in the existing parallelism recognition methods for the many-core processors.It could lead to serious load imbalance when the selected loop parallel dimension iteration number was small.To solve this problem,this paper proposed a multi-dimensional parallel recognition method for many-core processor.When it was difficult for the existing recognition methods to reach a better load balancing,this paper took a multi-dimensional parallel approach to the nested loops,and merged a task partition scheme after multi-dimensional parallel iteration space to reduce the impact of load imbalance on parallel efficiency of the program.It has been implemented in the automatic parallelization system developed by the research group,which can improve the parallel execution efficiency of some applications on many-core processor.
作者
李颖颖
庞建民
李雁冰
翟胜伟
Li Yingying;Pang Jianmin;Li Yanbing;Zhai Shengwei(State Key Laboratory of Mathematical Engineering&Advanced Computing,Zhengzhou 450002,China;Information Engineering University,Zhengzhou 450002,China;The 27th Research Institute of China Electronics Technology Group Corporation,Zhengzhou 450047,China)
出处
《计算机应用研究》
CSCD
北大核心
2018年第11期3311-3314,共4页
Application Research of Computers
基金
国家自然科学基金面上项目(61472447)
国家"863"计划资助项目(2014AA01A300)
国家"核高基"重大专项资助项目
关键词
多维并行识别
众核处理器
自动并行化
嵌套循环
multi-dimensional parallelism recognition
many-core processor
automatic parallelization
nested loop