Using the SMiLE monitoring infrastructure to detect and lower the inefficiency of parallel applications

Jie Tao, Wolfgang Karl, Martin Schulz

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

High computational demands are one of the main reasons for the use of parallel architectures like clusters of PCs. Many parallel programs, however, suffer from severe inefficiencies when executed on such a loosely coupled architecture for a variety of reasons. One of the most important is the frequent access to remote memories. In this article, we present a hybrid event-driven monitoring system which uses a hardware monitor to observe all of the underlying transactions on the network and to deliver information about the run-time behavior of parallel programs to tools for performance analysis and debugging. This monitoring system is targeted towards cluster architectures with NUMA characteristics.

Original languageEnglish
Title of host publicationHigh Performance Computing and Networking - 8th International Conference, HPCN Europe 2000, Proceedings
EditorsMarian Bubak, Hamideh Afsarmanesh, Bob Hertzberger, Roy Williams
PublisherSpringer Verlag
Pages270-279
Number of pages10
ISBN (Print)9783540675532
DOIs
StatePublished - 2000
Event8th International Conference on High Performance Computing and Networking, HPCNEurope 2000 - Amsterdam, Netherlands
Duration: 8 May 200010 May 2000

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1823
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th International Conference on High Performance Computing and Networking, HPCNEurope 2000
Country/TerritoryNetherlands
CityAmsterdam
Period8/05/0010/05/00

Fingerprint

Dive into the research topics of 'Using the SMiLE monitoring infrastructure to detect and lower the inefficiency of parallel applications'. Together they form a unique fingerprint.

Cite this