Parallel approximate factorization method for solving discrete elliptic equations

摘要

An iterative approximate factorization method is proposed for solving elliptic equations on a rectangular computational domain on distributed memory parallel computers. The method is based on the conjugate gradient algorithm with modified incomplete factorization pre- conditioning (MICCG(0)) and the use of a domain-decomposition-like ordering of unknowns. The rate of convergence and the efficiency of the proposed method are investigated.