期刊文献+

分子动力学模拟软件GROMACS向GPGPU移植初探 被引量:1

Research in porting MD simulation software GROMACS on GPGPU based on CUDA
原文传递
导出
摘要 GROMACS是著名分子动力学模拟软件之一,GPGPU技术能够使用图形处理器完成通用计算,是高性能计算的最新发展趋势。本文通过对Nvidia CUDA GPGPU编程模式以及GROMACS代码的研究分析,论证了将GROMACS移植到以GPGPU为计算核心的平台上的可行性,并通过算法和参数数据组织形式的改造实现了GROMACS中非键力计算函数从x86处理器向GPGPU的移植,获得显著的性能提升,计算加速比达到10倍以上,为GROMACS整体移植的实现奠定了基础。 GROMACS is a popular MD simulation software.GPGPU technology which uses GPU to perform general-purpose computation is a kind of new development trend in high performance computing.Based on analysis of Nvidia CUDA GPGPU programming mode and code in GROMACS,this paper attempted to create GPGPU version of GROMACS.Using GPGPU as computing core,the non-bonded force computing function in GROMACS has been implemented on GPGPU by transforming arithmetic and data structure.As a result,function performance is enhanced with speed 10 times faster which would support the feasibility for porting the entire GROMACS software on GPGPU.
出处 《计算机与应用化学》 CAS CSCD 北大核心 2010年第12期1603-1606,共4页 Computers and Applied Chemistry
基金 国家自然科学基金(20336040 20221603) 中国科学院创新课题基金(0822121119) 多相复杂系统国家重点实验室开放课题资助(2008-8)
关键词 分子动力学 GROMACS GPGPU CUDA molecular dynamics GROMACS GPGPU CUDA
  • 相关文献

参考文献22

  • 1Nvidia Tesla Computing Solutions.http://www.nvidia.com/obj ect/tesla_computing_solutions.html.
  • 2CUDA Zone.http://www.nvidia.com/object/cuda-home.html.
  • 3AMD FireStreamTM 9170:Industry's First GPU with Double-Precision Floating Point.
  • 4http://ati.amd.com/products/streamprocessor/specs.html.ATI Stream Software Development Kit.http://developer.amd.com/gpu/ATl Streams DK/Pages/de fanlt.aspx.
  • 5IBM Cell Broadband Engine technology.http://www-03.ibm.com/technology/cell/.
  • 6Cell Programming.http://www.ibm.eom/developerworks/cn/linux/1-cn-cellprogrammi ng/index.html.
  • 7Larrabee:Next-Generation Visual Computing Microarchitecture.http://www.intel.com/teclmology/visual/microarch.htm?iid=tech_ micro-cn+larrabee.
  • 8Larrabee (GPU)-Wikipedia,the free encyclopedia.http://en.wikipedia.org/wiki/Larrabee_(GPU).
  • 9Jiang Y.Pre-processing system and feasibility of a special-purpose computer cluster design for the particle methods.Master Degree Thesis.Beijing:China Academy of Sciences,2007.
  • 10Yen L,He M],Guo L.Summary of the establishment of large-scale panicle simulation parallel computing platform.Unpublicized Material.Beijing; Institute of Process Engineering,Chinese Academy of Sciences,2008.

同被引文献9

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部