期刊文献+

微机集群的并行蒙特卡罗模拟 被引量:5

The Parallel Monte-Carlo Simulation by the PC Clusters
下载PDF
导出
摘要 针对目前蒙特卡罗模拟并行计算的两个制约因素,提出了利于普及的完整解决方案。建立了算法、参数与MCNP一致的并行随机数发生器模块。参考Linux集群方案,提出了一种基于Windowsxp的pc机并行计算系统的实现方案,采用通用局域网络设置,并行环境为MPICH2实现,编程调试工具为MicrosoftDevelopStudio。该方案对并行系统中的计算机没有硬件和软件的一致性的严格要求,具有灵活的可扩展性,对于一台计算机的也可以虚拟并行,适合单机下的并行程序调试。最后给出了简单蒙特卡罗模拟实例进行演示。 A scheme suitable to the widely use of the parallel Monte-Carlo simulation was present. For resolve of the two main bottle nick in parallel Monte-Carlo simulation, referenced to the algorithms and constants used in MCNP code system, a parallel random number generator module was developed. Similar to Linux PC cluster scheme, the Windows xp PC cluster was made, computer nodes was connected by local network system, and use MPICH2 to setup the parallel environment, use Microsoft Develop Studio as programming tools. this system is much easier then those with Linux operation system, and may have different kind of the computers. So the system has better expansibility, and even only one computer can also debug the parallel computing programs. Finally a simple Monte-Carlo sample problem was present for the demonstration.
出处 《微电子学与计算机》 CSCD 北大核心 2006年第z1期210-212,共3页 Microelectronics & Computer
关键词 并行计算 MPICH2 蒙特卡罗模拟 随机数发生器 Parallel computing, MPICH2, Monte-carlo simulation, Random number generator
  • 相关文献

参考文献3

  • 1[1]J F Briesmeister.MCNP-A General Monte Carlo NParticle Transport Code[R].New Mexico:Los Alamos national Laboratory,2000:185~186
  • 2[2]F B Brown,Yasunobu Nagaya.The MCNP5 Random Number Generator[R],New Mexico:Los Alamos national Laboratory,2002:1~3
  • 3[3]W Gropp,E Lusk,D Buntinas,et.al.MPICH2 User's Guide[R],Chicago:Arognne National Laboratory,2005:1~26

同被引文献15

引证文献5

二级引证文献8

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部