Recent Enhancements To Pvm

Author:

Beguelin Adam1,Dongarra Jack2,Geist Al3,Manchek Robert4,Sunderam Vaidy5

Affiliation:

1. SCHOOL OF COMPUTER SCIENCE, CARNEGIE MELLON UNIVERSITY, PITTSBURGH, PENNSYLVANIA 15213, AND PITTSBURGH SUPERCOMPUTING CENTER

2. UNIVERSITY OF TENNESSEE, KNOXVILLE, TENNESSEE 37996-1301, OAK RIDGE NATIONAL LABORATORY BOX 2008, BUILDING 6012 OAK RIDGE, TENNESSEE 37831-6367

3. OAK RIDGE NATIONAL LABORATORY BOX 2008, BUILDING 6012 OAK RIDGE, TENNESSEE 37831-6367

4. UNIVERSITY OF TENNESSEE, KNOXVILLE, TENNESSEE 37996-1301

5. EMORY UNIVERSITY DEPARTMENT OF MATHEMATICS AND COMPUTER SCIENCE ATLANTA, GEORGIA 30322

Abstract

This paper presents new features of PVM, a popular standard for writing parallel programs that execute over networks of heterogeneous machines. Although PVM has become an important infrastructure for paral lel programmers, we continue to develop the system based both on user feedback and our own research interests. In this paper we present new communica tions routines and briefly characterize their perfor mance. We describe new extensible services that allow advanced users to customize certain aspects of the de fault PVM functionality. An overview of shared-mem ory PVM optimizations is presented. PVM's new trac ing facility and a graphical console that utilizes this capability are described. Finally, we discuss future ex tensions to PVM now under investigation.

Publisher

SAGE Publications

Subject

General Engineering,General Computer Science

Cited by 12 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Supporting data management on cluster grids;Future Generation Computer Systems;2008-02

2. Parallel computation on multilayer cluster grids;Concurrency and Computation: Practice and Experience;2007

3. Enabling PVM to exploit the SCTP protocol;Journal of Parallel and Distributed Computing;2006-11

4. Running Large-Scale Applications on Cluster Grids;The International Journal of High Performance Computing Applications;2005-05

5. Utilizing PVM in a Multidomain Clusters Environment;Recent Advances in Parallel Virtual Machine and Message Passing Interface;2005

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3