Algebraic laws for BSP programming

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

18 Scopus citations

Abstract

The Bulk-Synchronous Parallel (BSP) model of computation reflects the capabilities and costs of communication on a wide range of general-purpose parallel architectuers. Thus it allows general-purpose parallel software and hardware to be developed independently of one another; much as the von Neumann model provides the same facility for sequential computation. This paper presents a simple programming notation for shared-memory programming, based upon BSP constructs. The notation is defined formally by its effect upon process states. Algebraic laws are given which allow program derivation. A further set of laws allow transformation of finite programs to a normal form.

Original languageEnglish
Title of host publicationEuro-Par 1996 Parallel Processing - 2nd International Euro-Par Conference, Proceedings
EditorsLuc Bouge, Pierre Fraigniaud, Anne Mignotte, Yves Robert
PublisherSpringer Verlag
Pages359-368
Number of pages10
ISBN (Print)3540616276, 9783540616276
DOIs
StatePublished - 1996
Externally publishedYes
Event2nd International Euro-Par Conference on Parallel Processing, Euro-Par 1996 - Lyon, France
Duration: 26 Aug 199629 Aug 1996

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1124
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2nd International Euro-Par Conference on Parallel Processing, Euro-Par 1996
Country/TerritoryFrance
CityLyon
Period26/08/9629/08/96

Fingerprint

Dive into the research topics of 'Algebraic laws for BSP programming'. Together they form a unique fingerprint.

Cite this