Numerical experiments with monte carlo methods and SPAI preconditioner for solving system of linear equations

Abstract

In this paper we present the results of experiments comparing the performance of the mixed Monte Carlo algorithms and SPAI preconditener with BICGSTAB. The experiments are carried out on a Silicon Graphics ONYX2 machine. Based on our experiments, we con- clude that these techniques are comparable from the point of view of robustness and rates of convergence, with the Monte Carlo approach performing better for some general cases and SPAI approach performing better in case of very sparse matrices. © Springer-Verlag Berlin Heidelberg 2002.