Faculty Scholarly Dissemination Grants
Parallel implementations of FGMRES for solving large, sparse non-symmetric linear systems
Department
School of Computing and Information Systems
College
Padnos College of Engineering and Computing
Date Range
2012-2013
Disciplines
Engineering
Abstract
The Flexible Generalized Minimal Residual method (FGMRES) is an attractive iterative solver for non-symmetric systems of linear equations. This paper presents several methods for parallelizing FGMRES for a variety of architectures including multi-core CPU, Graphics Processing Units (GPU), and multi-GPU systems. The parallel implementations utilize OpenMP and CUDA kernels, and are organized according to thread scope. The linear systems employed in this study correspond to the discrete analogues of realistic three-dimensional convection-diffusion problems, and range in size to nearly 107 linear equations. All of the parallel implementations, particularly the novel hybrid approach, show a significant speedup over the sequential version. Theoretical insight and performance
Conference Name
Int. Conf. on Computational Science
Conference Location
Barcelona, Spain
ScholarWorks Citation
Wolffe, Greg; DeVries, Byron; Iannelli, Joe; Trefftz, Christian; and OHearn, Kurt, "Parallel implementations of FGMRES for solving large, sparse non-symmetric linear systems" (2013). Faculty Scholarly Dissemination Grants. 1004.
https://scholarworks.gvsu.edu/fsdg/1004
