Skip to Main content Skip to Navigation
Journal articles

Concurrent number cruncher - A GPU implementation of a general sparse linear solver

Luc Buatois 1 Guillaume Caumon 2 Bruno Lévy 1
1 ALICE - Geometry and Lighting
INRIA Lorraine, LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications
Abstract : A wide class of numerical methods needs to solve a linear system, where the matrix pattern of non-zero coefficients can be arbitrary. These problems can greatly benefit from highly multithreaded computational power and large memory bandwidth available on GPUs, especially since dedicated general purpose APIs such as CTM (AMD-ATI) and CUDA (NVIDIA) have appeared. CUDA even provides a BLAS implementation, but only for dense matrices (CuBLAS). Other existing linear solvers for the GPU are also limited by their internal matrix representation. This paper describes how to combine recent GPU programming techniques and new GPU dedicated APIs with high performance computing strategies (namely block compressed row storage, register blocking and vectorization), to implement a sparse general-purpose linear solver. Our implementation of the Jacobi-preconditioned Conjugate Gradient algorithm outperforms by up to a factor of 6.0x leading-edge CPU counterparts, making it attractive for applications which are content with single precision.
Complete list of metadatas

Cited literature [30 references]  Display  Hide  Download
Contributor : Nicolas Ray <>
Submitted on : Monday, October 20, 2008 - 9:35:10 AM
Last modification on : Thursday, May 28, 2020 - 3:10:07 PM
Long-term archiving on: : Monday, June 7, 2010 - 6:27:49 PM


Files produced by the author(s)




Luc Buatois, Guillaume Caumon, Bruno Lévy. Concurrent number cruncher - A GPU implementation of a general sparse linear solver. International Journal of Parallel, Emergent and Distributed Systems, Taylor & Francis, 2008, 24 (3), pp.205-223. ⟨10.1080/17445760802337010⟩. ⟨inria-00331906⟩



Record views


Files downloads