Iteration-merged Solver for Linear Banded Matrices Based on Domain Decomposition and MPI

LIU Zhaohui,SHU Jiwu,ZHENG Weimin
DOI: https://doi.org/10.3321/j.issn:1000-0054.2005.10.030
2005-01-01
Abstract:Parallel solvers for linear banded systems are often based on domain decomposition methods with two levels of iterations implemented with a message passing interface(MPI).However,excessive iterations,as well as excessive process communications, result in low parallel efficiency.This paper presents an efficient iterationmerged solver suitable for domain decomposition and MPI systems that minimizes the iterations and the process communications.A global acceleration algorithm merged two levels of iterations into one level to reduce the total number of iterations with a block-parallel technique to minimize the process communications on MPI systems caused by the global acceleration.Tests demonstrate that the parallel solver have the same number of iterations with the block-parallel technique reducing the process communication time for the global acceleration by half.
What problem does this paper attempt to address?