DaReL: A portable data redistribution library for distributed-memory machines

E. T. Kalns, Lionel M. Ni

Research output: Chapter in Book/Conference Proceeding/ReportConference Paper published in a bookpeer-review

Abstract

Run-time data redistribution can be performed between algorithm phases when a different data decomposition is expected to deliver increased performance for a subsequent phase of computation. Redistribution, however, represents increased program overhead as algorithm computation is interrupted while data are exchanged among processor memories. A number of data parallel Fortran languages support run-time data redistribution primitives. We propose a portable data redistribution library, DaReL, specifically for High Performance Fortran (HPF). DaReL supports multi-dimensional data redistribution for HPF's regular distribution patterns, BLOCK, CYCLIC, and ∗. Data exchange is performed with MPI primitives, facilitating DaReL's portability among distributed memory platforms that utilize the emerging message passing standard. We present an overview of DaReL's design and performance results on an IBM SP-1.

Original languageEnglish
Title of host publicationProceedings - Scalable Parallel Libraries Conference, SPLC 1994
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages78-87
Number of pages10
ISBN (Electronic)0818668954, 9780818668951
DOIs
Publication statusPublished - 1994
Externally publishedYes
Event1994 Scalable Parallel Libraries Conference, SPLC 1994 - Mississippi State, United States
Duration: 12 Oct 199414 Oct 1994

Publication series

NameProceedings - Scalable Parallel Libraries Conference, SPLC 1994

Conference

Conference1994 Scalable Parallel Libraries Conference, SPLC 1994
Country/TerritoryUnited States
CityMississippi State
Period12/10/9414/10/94

Bibliographical note

Publisher Copyright:
© 1994 IEEE.

Fingerprint

Dive into the research topics of 'DaReL: A portable data redistribution library for distributed-memory machines'. Together they form a unique fingerprint.

Cite this