A performance study of the RPE mechanism for PDES

Proceedings of International Workshop on Modeling, Analysis and Simulation of Computer and Telecommunication Systems Pub Date : 1994-01-31 DOI:10.1109/MASCOT.1994.284420

J. E. Butler, V. Wallentine

{"title":"A performance study of the RPE mechanism for PDES","authors":"J. E. Butler, V. Wallentine","doi":"10.1109/MASCOT.1994.284420","DOIUrl":null,"url":null,"abstract":"We present the latest implementation of our Remote Process Execution (RPE) mechanism for parallel discrete-event simulation (PDES), along with results of an empirical study of its performance. RPE is an optimistic time warp based mechanism, which allows logical processes on one processor to execute remotely on another when the former is lagging behind in the computation. The advantage is that processors running too far ahead need neither wait, nor run excessively optimistically: they can instead do some useful work for processors which are falling behind. The disadvantage of course, is the cost associated with running processes remotely. The experiments we have run so far show that performance can be significantly enhanced through RPE under the right conditions. In particular, as expected, the site of computation grains was found to be of primary importance. Though our tests were ran with relatively few processors, we suggest a few simple modifications which should allow the scheme to work well for large numbers of computing elements.<<ETX>>","PeriodicalId":288344,"journal":{"name":"Proceedings of International Workshop on Modeling, Analysis and Simulation of Computer and Telecommunication Systems","volume":"17 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1994-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of International Workshop on Modeling, Analysis and Simulation of Computer and Telecommunication Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MASCOT.1994.284420","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 3

Abstract

We present the latest implementation of our Remote Process Execution (RPE) mechanism for parallel discrete-event simulation (PDES), along with results of an empirical study of its performance. RPE is an optimistic time warp based mechanism, which allows logical processes on one processor to execute remotely on another when the former is lagging behind in the computation. The advantage is that processors running too far ahead need neither wait, nor run excessively optimistically: they can instead do some useful work for processors which are falling behind. The disadvantage of course, is the cost associated with running processes remotely. The experiments we have run so far show that performance can be significantly enhanced through RPE under the right conditions. In particular, as expected, the site of computation grains was found to be of primary importance. Though our tests were ran with relatively few processors, we suggest a few simple modifications which should allow the scheme to work well for large numbers of computing elements.<>

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

PDES RPE机制的性能研究

我们提出了用于并行离散事件模拟(PDES)的远程进程执行(RPE)机制的最新实现，以及对其性能的实证研究结果。RPE是一种基于乐观时间扭曲的机制，它允许一个处理器上的逻辑进程在另一个处理器上远程执行，当前者在计算中落后时。这样做的好处是，超前运行的处理器既不需要等待，也不需要过于乐观地运行:相反，它们可以为落后的处理器做一些有用的工作。当然，缺点是与远程运行进程相关的成本。到目前为止，我们运行的实验表明，在适当的条件下，通过RPE可以显著提高性能。特别是，正如预期的那样，计算粒的位置被发现是最重要的。虽然我们的测试是在相对较少的处理器上运行的，但我们建议进行一些简单的修改，使该方案能够很好地适用于大量的计算元素。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Proceedings of International Workshop on Modeling, Analysis and Simulation of Computer and Telecommunication Systems

自引率

0.00%

发文量

期刊最新文献

Time constrained message transmission in a LAN environment Approximate analysis of a multi-class open queueing network with class blocking and push-out Evaluating memory system performance of a large scale NUMA multiprocessor Efficient technique for performance analysis of locking protocols A maximum entropy analysis of the single server queue