Iterative Solution on GPU of Linear Systems Arising from the A-V Edge-FEA of Time-Harmonic Electromagnetic Phenomena

Author

Camargos, Ana F. P. ; Silva, Viviane Cristine ; Guichon, Jean-M ; Meunier, Gerard

Author_Institution

Inst. Fed. de Minas Gerais, Formiga, Brazil

fYear

2014

fDate

12-14 Feb. 2014

Firstpage

365

Lastpage

371

Abstract

We present a performance analysis of a parallel implementation to both preconditioned Conjugate Gradient and preconditioned Bi-conjugate Gradient solvers using graphic processing units with CUDA programming model. The solvers were optimized for the solution of sparse systems of equations arising from Finite Element Analysis of electromagnetic phenomena involved in the diffusion of underground currents under time-harmonic current excitation. We used a shifted Incomplete Cholesky factorization as preconditioner. Results show a significant speedup by using the GPU compared to a serial CPU implementation.

Keywords

computational electromagnetics; conjugate gradient methods; eddy currents; finite element analysis; graphics processing units; iterative methods; linear systems; parallel architectures; A-V edge-FEA; GPU; eddy currents; finite element analysis; graphic processing unit; iterative solution; linear systems; preconditioned biconjugate gradient solver; preconditioned conjugate gradient solver; shifted incomplete Cholesky factorization; sparse systems; time-harmonic current excitation; time-harmonic electromagnetic phenomena; underground currents; Equations; Finite element analysis; Graphics processing units; Iterative methods; Linear systems; Mathematical model; Sparse matrices; Finite Elements; Graphic Processing Unit; Incomplete Factorization; Preconditioner;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel, Distributed and Network-Based Processing (PDP), 2014 22nd Euromicro International Conference on

Conference_Location

Torino

ISSN

1066-6192

Type

conf

DOI

10.1109/PDP.2014.95

Filename

6787300