The performance of the CUDA implementations of HABC and CFS-PML ABCs on GPU hardware

Author

Almeer, Hanan ; Costen, Fumie ; Berenger, Jean-Pierre ; Himeno, Ryutaro ; Yokota, Hideo

Author_Institution

Sch. of Electr. & Electron. Eng., Univ. of Manchester, Manchester, UK

fYear

2014

fDate

6-11 April 2014

Firstpage

1161

Lastpage

1164

Abstract

The General Purpose computing on Graphics Processing Units (GPGPU) is used to accelerate the FDTD method in an effective way. This paper presents the performance of the Huygens Absorbing Boundary Conditions (HABC) implementation on GPU and the Complex Frequency-Shifted Perfectly Matched Layer (CFS-PML) Absorbing Boundary Conditions (ABCs) implementation on the GPU. A comparison between the implementations of the two ABCs on Tesla architecture GPUs is proposed in this paper.

Keywords

finite difference time-domain analysis; general purpose computers; graphics processing units; parallel architectures; CFS-PML ABC; CUDA implementations; FDTD method; GPGPU; GPU hardware; HABC; Huygens absorbing boundary conditions; complex frequency shifted perfectly matched layer absorbing boundary conditions; general purpose computing on graphics processing units; Computer architecture; Equations; Finite difference methods; Graphics processing units; Instruction sets; Mathematical model; Time-domain analysis; ABC; FDTD; GPU; HPC;

fLanguage

English

Publisher

ieee

Conference_Titel

Antennas and Propagation (EuCAP), 2014 8th European Conference on

Conference_Location

The Hague

Type

conf

DOI

10.1109/EuCAP.2014.6901978

Filename

6901978