Massively parallel two-dimensional TLM algorithm on graphics processing units

Author

Rossi, Filippo V. ; So, Poman P M ; Fichtner, Nikolaus ; Russer, Peter

Author_Institution

Dept. ECE, University of Victoria, British Columbia, Canada

fYear

2008

fDate

15-20 June 2008

Firstpage

153

Lastpage

156

Abstract

Recent advances in computing technology has brought massively parallel computing power to desktop PCs. As multi-core processor technology becomes mature, a new front in parallel technology based on graphics processors has emerged. A massively parallel 2D-TLM algorithm for NVIDIA advanced graphics processors has been developed. The proposed parallel computing paradigm can be adopted straightforwardly to accelerate time-domain electromagnetic field modeling programs.

Keywords

coprocessors; parallel processing; NVIDIA advanced graphics processors; graphics processing units; massively parallel computing power; massively parallel two-dimensional TLM algorithm; multi-core processor technology; time-domain electromagnetic field modeling programs; Acceleration; Central Processing Unit; Electromagnetic scattering; Finite difference methods; Graphics; Kernel; Parallel processing; Synchronization; Time domain analysis; Yarn; FDTD; GPU; Parallel Algorithms; SIMD; Stream Computing; TLM; Time Domain Computational Electromagnetics;

fLanguage

English

Publisher

ieee

Conference_Titel

Microwave Symposium Digest, 2008 IEEE MTT-S International

Conference_Location

Atlanta, GA

ISSN

0149-645X

Print_ISBN

978-1-4244-1780-3

Electronic_ISBN

0149-645X

Type

conf

DOI

10.1109/MWSYM.2008.4633126

Filename

4633126