DocumentCode
2540476
Title
On optimizing collective communication
Author
Chan, Ernie W. ; Heimlich, Marcel F. ; Purkayastha, Avi ; Van de Geijn, Robert A.
Author_Institution
Univ. of Texas, Austin, TX, USA
fYear
2004
fDate
20-23 Sept. 2004
Firstpage
145
Lastpage
155
Abstract
We discuss issues related to the high-performance implementation of collective communications operations on distributed-memory computer architectures. Using a combination of known techniques (many of which were first proposed in the 1980s and early 1990s) along with careful exploitation of communication modes supported by MPI, we have developed implementations that have improved performance in most situations compared to those currently supported by public domain implementations of MPI such as MPICH. Performance results from a large Intel Pentium 4 (R) processor cluster are included.
Keywords
computer architecture; distributed memory systems; message passing; optimisation; MPI; collective communication optimization; communication modes; distributed-memory computer architectures; public domain implementations; Computational efficiency; Computational modeling; Computer architecture; Concurrent computing; Costs; Distributed computing; High performance computing; Network topology; Protocols; Routing;
fLanguage
English
Publisher
ieee
Conference_Titel
Cluster Computing, 2004 IEEE International Conference on
ISSN
1552-5244
Print_ISBN
0-7803-8694-9
Type
conf
DOI
10.1109/CLUSTR.2004.1392612
Filename
1392612
Link To Document