Abstract
The performance of collective communication is critical to the overall system performance. In general, the performance of the collective communication is dependent not only on the underlying hardware, but also its implementation. To evaluate the performance of collective communication accurately, identifying the most representative metrics and using correct measurement techniques are two important issues that have to be considered. This paper focuses on the measurement techniques for collective communication services. The proposed techniques can provide an accurate evaluation of the completion time without requiring a global clock and without having to know the detailed implementations of collective communication services. Experimental results conducted on the IBM/SP at Argonne National Laboratory are presented.
| Original language | English |
|---|---|
| Title of host publication | Communication and Architectural Support for Network-Based Parallel Computing - 1st International Workshop, CANPC 1997, Proceedings |
| Editors | Craig B. Stunkel, Dhabaleswar K. Panda |
| Publisher | Springer Verlag |
| Pages | 212-226 |
| Number of pages | 15 |
| ISBN (Print) | 3540625739, 9783540625735 |
| DOIs | |
| Publication status | Published - 1997 |
| Externally published | Yes |
| Event | 1st Workshop on Communication and Architectural Support for Network-based Parallel Computing, CANPC 1997 - San Antonio, United States Duration: 1 Feb 1997 → 2 Feb 1997 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 1199 |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Conference
| Conference | 1st Workshop on Communication and Architectural Support for Network-based Parallel Computing, CANPC 1997 |
|---|---|
| Country/Territory | United States |
| City | San Antonio |
| Period | 1/02/97 → 2/02/97 |
Bibliographical note
Publisher Copyright:© Springer-Verlag Berlin Heidelberg 1997.