面向科研院校服务的高性能计算集群平台的设计与实现

打开文本图片集
摘 要:本研究阐述了高性能计算集群平台的特点,然后介绍了平台的硬件和软件组成,并简要地阐述了平台的构建过程,最后通过一个测试案例探讨了利用平台进行作业调度的流程和方法,希望可以为企业高性能计算运维人员提供一些借鉴和参考。
关键词:高性能计算;集群部署;高校科学计算
中图分类号:TP38 文献标志码:A 文章编号:1003-5168(2022)3-0026-04
DOI:10.19968/j.cnki.hnkj.1003-5168.2022.03.006
Design and Implementation of High Performance Computing Cluster Platform for Scientific Research Institutions
TU Jinlong
(Network and Information Technology Center of Jiangxi University of Chinese Medicine,
Nanchang 330004,China)
Abstract:This paper describes the characteristics of high-performance computing cluster platform, then introduces the hardware and software composition of the platform, and briefly describes the construction process of the platform. Finally, through a test case, this paper discusses the process and method of job scheduling using the platform, hoping to provide some reference for enterprise high-performance computing operation and maintenance personnel.
Keywords:high performance computing; cluster deployment; scientific computing in colleges and universities
0 引言
高性能计算(High Performance Computing,HPC)也被称为数据分析的超级计算机,是一套计算性能强大、数据传输率超高、具有大规模存储空间和完整软件的系统,通常包含计算处理单元、并行存储单元、网络单元、管理调度软件单元,主要采用集群架构,通过网络将大量的服务器连接起来,让所有服务器协调工作来完成一系列计算任务[1]。(剩余4776字)