PC CLUSTER地震资料处理系统集成与应用
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
联网微机、并行中间件、实时监视和自动部署是构成PC Cluster的四个基本要素,本文探讨了如何根据地震资料处理的技术特点集成PC Cluster地震资料处理系统,提出了利用开放源带码项目进行软件集成的概念。通过应用实例对比分析了PC Cluster地震资料并行处理效率,得出网络I/O与扰动是影响PCCluster性能的关键因素并提出了解决办法。
PC Cluster served as a high performance computing platform of seismic data processing must have four elements: networked PC,parallel midware,real time monitoring and automatic distribution.How to integrate PC cluster seismic data processing system based on seismic data process is discussed in this paper.It is found that Network I/O and disturbing are the main bottlenecks of PC Cluster with the analysis of some examples about seismic parallel data processing,and the solution to solve these problems is given.
引文
[1]Wilkison B,Allen M.Parallel Programming:Tech-nique and Applications Using Networked Workstationand Parallel Computers[M].Prentice Hall Inc.,1999.
    [2]D J Becker,T Sterling,D Savarese,et al.Beowulf:A Parallel Workstation for Scientific Computation[A].in:Proceedings of the 9th International Confer-ence on Parallel Processing[C].1995.
    [3]Anderson T E,D E Culler,D Patterson.A Case forNOW(Networks of Workstations)[J].IEEE Mirco,1995,15(1):54-64.
    [4]赖能和.PC CLUSTER系统集成技术与并行效率分析[J].石油地球物理,2003,38(增刊):103-107.
    [5]Pacheco P.Parallel Programming with MPI,MorganKaufman[M].San Francisco,1997.
    [6]都志辉.高性能计算并行编程技术MPI并行程序设计[M].清华大学出版社,2001.
    [7]A Geist,A Beguelin,J Dongarra,et al.PVM:A Us-er'Guide and Tutorial for Networked Parallel Compu-ting[M].MIT Press,1994.
    [8]M L Massie,B N Chun,D E Culler.The GangliaDistributed Monitoring System:Design,Implementa-tion,and Experience[J].Parallel Computing,2004,30(7):817-840.

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心