GPU Acceleration of Finite Difference Schemes Used in Coupled Electromagnetic/Thermal Field Simulations

Author

Richter, Chris ; Schops, Sebastian ; Clemens, Markus

Author_Institution

Bergische Univ. Wuppertal, Wuppertal, Germany

Volume

49

Issue

5

fYear

2013

fDate

May-13

Firstpage

1649

Lastpage

1652

Abstract

The solution procedure of coupled electromagnetic-/thermal-simulations with high resolution requires efficient solvers. High performance computing libraries and languages like Nvidia´s CUDA help in unlocking the massively parallel capabilities of GPUs to accelerate calculations. They reduce the time needed to solve real world problems. In this paper, the speed-up is discussed, which is obtained by using GPUs for coupled time domain simulations with finite difference schemes. A tailor-made implementation of the time consuming sparse matrix vector multiplication is shown to have advantages over standard CUDA-libraries like cuSparse.

Keywords

computational electromagnetics; electromagnetic field theory; finite difference time-domain analysis; graphics processing units; parallel architectures; sparse matrices; CUDA-libraries; GPU acceleration; Nvidia CUDA; coupled electromagnetic-thermal field simulations; coupled time domain simulations; cuSparse; efficient solvers; finite difference schemes; high performance computing languages; high performance computing libraries; solution procedure; speed-up; tailor-made implementation; time consuming sparse matrix vector multiplication; Bioheat; CUDA; GPU; SpMV; conjugate gradients; finite differences; multiphysics;

fLanguage

English

Journal_Title

Magnetics, IEEE Transactions on

Publisher

ieee

ISSN

0018-9464

Type

jour

DOI

10.1109/TMAG.2013.2238662

Filename

6514795