Abstract | ||
---|---|---|
Evaluating the performance of parallel processors and supercomputers is difficult because so many factors contribute to speed. The architecture, compiler, and algorithms used are all important, as are their interactions. In the past, it has been common practice to attempt to characterize each machine's performance by just one number. The complexity of the situation indicates that more work is needed to discover how best to characterize machines. An approach for the future is sketched in this paper. |
Year | Venue | Keywords |
---|---|---|
1987 | Proceedings of the 1st International Conference on Supercomputing | supercomputing performance evaluation plan |
Field | DocType | Volume |
Architecture,Supercomputer,Computer science,Parallel processing,Parallel computing,Compiler,Turnaround time | Conference | 297 |
ISSN | ISBN | Citations |
0302-9743 | 0-387-18991-2 | 4 |
PageRank | References | Authors |
15.66 | 5 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
D. J. Kuck | 1 | 4 | 15.66 |
A. H. Sameh | 2 | 562 | 212.82 |