CUDA架构下分子动力学模拟的高速实现

 陈捷捷1,王彦浩2,刘丹3

(1.中国舰船研究设计中心,湖北 武汉 430064;2.清华大学 计算机科学与技术系,北京 100083;3.武汉第二船舶设计研究所,湖北 武汉 430064)
摘要:统一计算设备架构(CUDA)的提出和图形处理器(GPU)的并行处理能力和数据传输能力,使得基于CUDA架构的GPU高性能计算迅速成为一个研究热点。针对含有大规模分子动力学模拟效率低下问题,提出了一种CUDA架构下分子动力学模拟的新方法,给出了分子动力学模拟的CUDA算法流程,并利用CUDA C实现了并行算法。实验结果表明,与CPU处理速度相比,GPU可以提高速度200倍左右。
关键词:图形处理器;统一计算设备架构;分子动力学模拟
中图分类号:TP391.9          文献标志码:A           文章编号:1006-0316 (2013) 12-0073-04
High speed molecular dynamics simulation approach based on CUDA
CHEN Jie-jie1,WANG Yan-hao2,LIU Dan3
( 1.China Ship Development and Design Center, Wuhan 430064, China; 2.Tsinghua University Department of Computer Science and Technology, Beijing 100083, China; 
3.Wuhan Second Ship Design and Research Institute, Wuhan 430064, China )
Abstract:The Compute Unified Device Architecture (CUDA) and the rapidly promoting ability of data-parallel processing and data transferring of Graphic Processing Units (GPU) make an attractive area for high performance computation based on CUDA. As the efficiency of large scale molecular dynamics simulation is relative low, a new approach to the molecular dynamics simulations based on CUDA is proposed. The CUDA-based algorithm of molecular dynamics simulation is give. We have used the CUDA C to implement parallel algorithm. The results indicated that the performance of the GPU is compared to their main processor counterpart. We achieve speedups of up to near 200 folds.
Key words:GPU;CUDA;molecular dynamics simulation
 
———————————————
收稿日期:2013-04-13
作者简介:陈捷捷(1982-),江苏启东人,硕士,工程师,主要研究方向为虚拟现实与仿真。
 

 

设为首页  |  加入收藏    |   免责条款
《机械》杂志版权所有     Copyright©2008-2012 Jixiezazhi.com All Rights Reserved 

  电话:028-85925070    传真:028-85925073    E-mail:jixie@vip.163.com

地址:四川省成都锦江工业开发区墨香路48号   邮编:610063

蜀ICP备08103512号

Powered by PageAdmin CMS