ACCELERATION OF MOLECULAR MODELING APPLICATIONS WITH GRAPHICS PROCESSORS

使用图形处理器加速分子建模应用

• 批准号: 8363650
• 负责人: John H Stone
• 金额: $ 3.32万
• 依托单位: UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN
• 依托单位国家: 美国
• 财政年份: 2011
• 资助国家: 美国
• 起止时间: 2011-08-01 至 2012-09-09
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8363650
• 关键词: Acceleration; Architecture; Bioinformatics; Computer software; Funding; Generations; Grant; Individual; Modeling; Molecular Models; National Center for Research Resources; Performance; Principal Investigator; Process; Research; Research Infrastructure; Resources; Source; Time; United States National Institutes of Health; cost; molecular modeling; operation

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. Over the past several years, the hardware and software architecture of graphics processing units (GPUs) has evolved to the point that they can now be used for general purpose scientific computations. State-of-the-art graphics processors include hundreds of individual arithmetic units and can perform up to 500 billion floating point operations per second, a level of performance far above that available with current generation CPU cores. The Resource has implemented several GPU-accelerated computational kernels for key molecular modeling tasks which achieve performance levels of ten to one hundred times that of traditional CPU implementations (http://www.ks.uiuc.edu/Research/gpu/).

该副本是利用资源的众多研究子项目之一 由NIH/NCRR资助的中心赠款提供。对该子弹的主要支持 而且，副投影的主要研究员可能是其他来源提供的 包括其他NIH来源。 列出的总费用可能 代表subproject使用的中心基础架构的估计量， NCRR赠款不直接向子弹或副本人员提供的直接资金。 在过去的几年中，图形处理的硬件和软件体系结构 单位（GPU）已演变为现在可以用于通用目的的地步 科学计算。 最先进的图形处理器包括数百个个人 算术单元，每秒最多可以执行5000亿个浮点操作 远高于当前CPU内核可用的性能。 这 资源已经实施了几个GPU加速计算内核，用于关键分子 建模任务达到10到一百倍的绩效水平 传统CPU实现（http://www.ks.uiuc.edu/research/gpu/）。