首页 | 本学科首页   官方微博 | 高级检索  
     检索      

耗散粒子动力学GPU并行计算研究
引用本文:林晨森,陈硕,李启良,杨志刚.耗散粒子动力学GPU并行计算研究[J].物理学报,2014,63(10):104702-104702.
作者姓名:林晨森  陈硕  李启良  杨志刚
作者单位:1. 同济大学航空航天与力学学院, 上海 200092;2. 同济大学上海地面交通工具风洞中心, 上海 201804
基金项目:中央高等学校基本科研基金(批准号:125065);国家自然科学基金(批准号:51276130,10872152);教育部高等学校博士学科点专项科研基金(批准号:20120072110037)资助的课题~~
摘    要:研究了耗散粒子动力学基于计算统一设备架构的图形处理器(GPU)并行计算的实施.对其中涉及的算法映射模型、Cell-List法数组的并行化更新、随机数生成、存储器访问优化、负载平衡等进行了详细的讨论.进一步模拟了Poiseuille流动和突扩突缩流动,从而验证了GPU计算结果的正确性.计算结果表明,相对于基于中央处理器的串行计算,在耗散粒子动力学中实施GPU并行计算可以获得约20倍的加速比.

关 键 词:耗散粒子动力学  计算统一设备架构  图形处理器  并行计算
收稿时间:2013-12-10

Accelerating dissipative particle dynamics with graphic processing unit
Lin Chen-Sen,Chen Shuo,Li Qi-Liang,Yang Zhi-Gang.Accelerating dissipative particle dynamics with graphic processing unit[J].Acta Physica Sinica,2014,63(10):104702-104702.
Authors:Lin Chen-Sen  Chen Shuo  Li Qi-Liang  Yang Zhi-Gang
Abstract:In this paper, the graphic processing unit (GPU) parallel computing of dissipative particle dynamics (DPD) based on compute unified device architecture is carried out. Some issues involved, such as thread mapping, parallel cell-list array updating, generating pseudo-random number on GPU, memory access optimization and loading balancing are discussed in detail. Furthermore, Poiseuille flow and suddenly contracting and expanding flow are simulated to verify the correctness of GPU parallel computing. The results of GPU parallel computing of DPD show that the speedup ratio is about 20 times compared with central processing unit serial computing.
Keywords: dissipative particle dynamics compute unified device architecture graphic processing unit parallel computing
Keywords:dissipative particle dynamics  compute unified device architecture  graphic processing unit  parallel computing
本文献已被 CNKI 等数据库收录!
点击此处可从《物理学报》浏览原始摘要信息
点击此处可从《物理学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号