DMCA
An algorithm-by-blocks for SuperMatrix band Cholesky factorization
Cached
Download Links
- [www.cs.utexas.edu]
- [www.cs.utexas.edu]
- [vecpar.fe.up.pt]
- DBLP
Other Repositories/Bibliography
| Venue: | In VECPAR ’08: Proceedings of the Eighth International Meeting on High Performance Computing for Computational Science |
| Citations: | 3 - 3 self |
BibTeX
@INPROCEEDINGS{Quintana-ortí_analgorithm-by-blocks,
author = {Gregorio Quintana-ortí and Enrique S. Quintana-ortí and Alfredo Remón and Robert A. Van De Geijn},
title = {An algorithm-by-blocks for SuperMatrix band Cholesky factorization},
booktitle = {In VECPAR ’08: Proceedings of the Eighth International Meeting on High Performance Computing for Computational Science},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Abstract. We pursue the scalable parallel implementation of the factorization of band matrices with medium to large bandwidth targeting SMP and multi-core architectures. Our approach decomposes the computation into a large number of fine-grained operations exposing a higher degree of parallelism. The SuperMatrix run-time system allows an out-of-order scheduling of operations that is transparent to the programmer. Experimental results for the Cholesky factorization of band matrices on two parallel platforms with sixteen processors demonstrate the scalability of the solution.
Keyphrases
supermatrix band cholesky factorization band matrix multi-core architecture fine-grained operation large bandwidth sixteen processor parallel platform supermatrix run-time system large number cholesky factorization experimental result scalable parallel implementation out-of-order scheduling
