MRI: Acquisition of High Performance Clusters for Effective Parallel Computing in Computational Science Research and Education

voigt, Robert G.Simha， RahulCollege of William & Mary mri：在计算科学研究和教育中获得高效并行计算的高性能集群PIs申请资金购买和运行高速计算机系统网络，这将补充国家在William & Mary学院计算科学方面的投资，并帮助推动重要的相关研究和教育活动。该设备将被用作共同研究焦点的核心：有效地使用并行系统进行科学计算。pi寻求能够平衡其活动的许多需求的硬件。所选择的体系结构由三个相互连接的子系统组成，每个子系统都是工作站或PC的集群。其中两个是基于pentium的16处理器和32处理器的beowulf系统，采用不同的互连网络，第三个是四个4处理器SUN多处理器的组合。

EIA-9977030Voigt, Robert G.Simha, RahulCollege of William & MaryMRI: Acquisition of High Performance Clusters for Effective Parallel Computing in Computational Science Research and EducationThe PIs request funding to purchase and operate a high speed network of computer systems that will complement state investment in computational science at College of William and Mary and help drive significant, related research and educational activities. The equipment will be used as the centerpiece of the common research focus: the effective use of parallel systems for scientific computation. The PIs seek hardware that balances the many needs of their activities. The selected architecture consists of three interconnected subsystems each of which is a cluster of workstations or PC's. Two of these are 16-processor and 32-procesor Pentium-based Beowulf-like systems built with different interconnection networks, and the third is a combination of four 4-processor SUN multiprocessors.