Acquisition of a High Performance Parallel Computer: Multi-Disciplinary Need to Advance Computing in Engineering

购买高性能并行计算机：推进工程计算的多学科需求

• 批准号: 9977568
• 负责人: Bakhtier Farouk
• 金额: $ 20万
• 依托单位: Drexel University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1999
• 资助国家: 美国
• 起止时间: 1999-09-01 至 2003-02-28
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9977568&HistoricalAwards=false
• 关键词: Acquisition; Performance; Parallel; Computer; Multi

This project provides support for acquisition of high-end parallel processing computer for the college of Engineering at Drexel University.This parallel computing platform will support cross-disciplinary research needs in performing challenging numerical computations on single workstations or overloaded network servers. The parallel computing server will provide a shared memory computing environment with low-latency, high bandwidth communication between processors. It has many advantages including: a stable technology, minimum required support personnel, scalability, an extensive development environment, software tools for converting programs to parallel, and low cost relative to massivey parallel systems. It would enable quick startup of the parallel machine and rapid transition of current computational models to the parallel platform.The high-performance parallel server will benefit research programs in computational fluid dynamics, computational mechanics, materials processing, microstructure development, dynamic optimization, widespread fatigue damage, communications, trafficking, computer architecture modeling, and population dynamics. Researchers can solve larger problems and can solve computationally intensive problems more rapidly.This is an equipment grant supported under MRI program.

本项目为Drexel大学工程学院购置高端并行处理计算机提供支持。该并行计算平台将支持跨学科研究的需要，在单个工作站或过载的网络服务器上执行具有挑战性的数值计算。 并行计算服务器将提供共享存储器计算环境，具有处理器之间的低延迟、高带宽通信。 它有许多优点，包括：一个稳定的技术，所需的最低限度的支持人员，可扩展性，广泛的开发环境，软件工具，转换程序并行，和低成本相对于大量的并行系统。 它将使并行机的快速启动和当前计算模型的快速过渡到并行平台，高性能并行服务器将有利于计算流体动力学，计算力学，材料加工，微观结构开发，动态优化，广泛的疲劳损伤，通信，交通，计算机体系结构建模和人口动力学的研究计划。 研究人员可以解决更大的问题，可以更快地解决计算密集型的问题。这是一个在MRI计划下支持的设备补助金。