A novel hardware acceleration technique for high performance parallel FDTD method

Author

Wenhua Yu ; Xiaoling Yang ; Yongjun Liu

Author_Institution

2COMU, State College, PA, USA

fYear

2011

fDate

8-10 Aug. 2011

Firstpage

149

Lastpage

153

Abstract

In this paper, we introduce one novel hardware acceleration technique based on a vector unit built in a regular CPU for high performance electromagnetic simulation tools. We investigate the performance of parallel conformal FDTD method on the Intel and AMD processors accelerated by the Vector Arithmetic Logic Unit (VALU), high performance cluster, and Graphics Processing Unit (GPU). The FDTD method that is parallel in nature is one of the most popular numerical methods to simulate various electromagnetic problems and phenomenon. Several examples are employed to demonstrate the engineering applications of parallel conformal FDTD method based on the VALU acceleration.

Keywords

computational electromagnetics; computer graphic equipment; coprocessors; finite difference time-domain analysis; floating point arithmetic; multiprocessing systems; vector processor systems; AMD processors; FDTD method; Intel processors; VALU; electromagnetic problems; electromagnetic simulation tools; graphics processing unit; hardware acceleration technique; numerical methods; parallel conformal method; regular CPU; vector arithmetic logic unit; Acceleration; Computational modeling; Computers; Finite difference methods; Graphics processing unit; Time domain analysis; FDTD; SSE; VALU; hardware acceleration;

fLanguage

English

Publisher

ieee

Conference_Titel

Electromagnetics, Applications and Student Innovation (iWEM), 2011 IEEE International Workshop on

Conference_Location

Taipei

Print_ISBN

978-1-61284-462-6

Electronic_ISBN

978-1-61284-461-9

Type

conf

DOI

10.1109/iWEM.2011.6021490

Filename

6021490