High Performance Complete and Incomplete Factorizations for Very Large Sparse Systems by using Scotch and PaStiX softwares - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2004

High Performance Complete and Incomplete Factorizations for Very Large Sparse Systems by using Scotch and PaStiX softwares

Abstract

PaStiX is a scientific library that provides a high performance direct supernodal solver for very large sparse linear systems. It relies on a block factorization based on an hybrid ordering (Nested Dissection + Halo Approximate Minimum Degree) obtained using the Scotch library. Efficient static scheduling and memory management are used to solve irregular problems with more of 25 millions unknowns on clusters of SMP nodes. In order to solve larger 3D problems, we apply these blockwise algorithms to compute robust and efficient parallel ILU preconditioners.

Keywords

No file

Dates and versions

inria-00346622 , version 1 (11-12-2008)

Identifiers

  • HAL Id : inria-00346622 , version 1

Cite

Pascal Hénon, Francois Pellegrini, Pierre Ramet, Jean Roman, Yousef Saad. High Performance Complete and Incomplete Factorizations for Very Large Sparse Systems by using Scotch and PaStiX softwares. Eleventh SIAM Conference on Parallel Processing for Scientific Computing, 2004, San Francisco, United States. ⟨inria-00346622⟩
73 View
0 Download

Share

Gmail Facebook X LinkedIn More