摘要
本文试图说明以下观点:越来越多的高性能应用要求利用地理上分布的、各式各样的计算和数据资源。这些应用希望能够通过高速网络将地理上分布、异构的各种高性能计算机、数据服务器、大型检索存储系统和可视化、虚拟现实系统等连接并集成起来,形成一个网络虚拟计算机(称为元计算机),来实现应用计算问题。这种元计算实质上就是基于高速网络的广域高性能并行与分布式计算。
The paper attempts to declare the following viewpoints:Emerging highperformance applications require the ability to exploit diverse,geographically distributed resources.These applications use highspeed networks to integrate supercomputers,large databases,archival storage devices,advanced visualization devices,and/or scientific instruments to form networked virtual supercomputers or metacomputers.The metacomputing,in fact,is widearea highperformance parallel and distributed computing.
出处
《计算机工程与科学》
CSCD
1998年第3期1-5,共5页
Computer Engineering & Science
关键词
广域
元计算
计算机网络
并行计算
分布式计算
parallel and distributed computing,widearea computing,metacomputing.