Sparse-Matrix-CG-Solver in CUDA
Institute of Computer Science II, Rheinische Friedrich-Wilhelms-Universitat Bonn, Bonn / Germany
Central European Seminar on Computer Graphics (CESCG), 2011
@article{michels2011sparse,
title={Sparse-Matrix-CG-Solver in CUDA},
author={Michels, D.},
year={2011}
}
This paper describes the implementation of a parallelized conjugate gradient solver for linear equation systems using CUDA-C. Given a real, symmetric and positive definite coefficient matrix and a right-hand side, the parallized cg-solver is able to find a solution for that system by exploiting the massive compute power of todays GPUs. Comparing sequential CPU implementations and that algorithm we achieve a speed up from 4 to 7 depending on the dimension of the coefficient matrix. Additionally the concept of preconditioners to decrease the time to find a solution is evaluated using the SSOR method. In the end additional suggestions are provided to further increase the speed of the presented CUDA cg-solver.
December 7, 2011 by hgpu
