An efficient multi-algorithms sparse linear solver for GPUs

Thomas Jost 1 Sylvain Contassot-Vivier 1 Stéphane Vialle 1, 2
1 ALGORILLE - Algorithms for the Grid
INRIA Lorraine, LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications
Abstract : We present a new sparse linear solver for GPUs. It is designed to work with structured sparse matrices where all the non-zeros are on a few diagonals. Several iterative algorithms are implemented, both on CPU and GPU. The GPU code is designed to be fast yet simple to read and understand. It aims to be as accurate as possible, even on chips that do not support double-precision floating-point arithmetic. Several benchmarks show that GPU algorithms are much faster than their CPU counterpart while their accuracy is satisfying.
Complete list of metadatas

https://hal-supelec.archives-ouvertes.fr/hal-00485963
Contributor : Sébastien van Luchene <>
Submitted on : Monday, May 24, 2010 - 2:45:32 PM
Last modification on : Thursday, March 29, 2018 - 11:06:04 AM

Identifiers

Collections

Citation

Thomas Jost, Sylvain Contassot-Vivier, Stéphane Vialle. An efficient multi-algorithms sparse linear solver for GPUs. B. Chapman, F. Desprez, G.R. Joubert, A. Lichnewsky, F. Peters and T. Priol. Parallel Computing: From Multicores and GPU's to Petascale (Volume 19), 19, IOS Press, pp.546-553, 2010, Advances in Parallel Computing, ⟨10.3233/978-1-60750-530-3-546⟩. ⟨hal-00485963⟩

Share

Metrics

Record views

458