Utilization of CUDA-OpenGL interoperability to display electromagnetic fields calculated by FDTD

Author

Demir, Veysel ; Elsherbeni, Atef Z.

Author_Institution

Northern Illinois Univ., DeKalb, IL, USA

fYear

2011

fDate

10-13 Aug. 2011

Firstpage

95

Lastpage

98

Abstract

In this contribution, a compute unified device architecture (CUDA) implementation of a two-dimensional finite-difference time-domain (FDTD) program will be presented along with the OpenGL interoperability to display electromagnetic fields. CUDA, which runs on a graphics processing unit (GPU) card, is used for electromagnetic field data generation and image manipulation, while OpenGL is used to draw field distribution on the screen. Since CUDA and OpenGL both run on GPU and share data through common memory in the framebuffer the CUDA-OpenGL interoperability is very fast. Step by step details of implementation of this interoperability are demonstrated.

Keywords

computational electromagnetics; computer graphic equipment; coprocessors; electromagnetic fields; finite difference time-domain analysis; open systems; CUDA-OpenGL interoperability; FDTD; GPU; compute unified device architecture; electromagnetic field data generation; graphic processing unit; image manipulation; two-dimensional finite-difference time-domain program; Arrays; Electromagnetic fields; Finite difference methods; Graphics processing unit; Kernel; Programming; Time domain analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Electromagnetics International Workshop (CEM), 2011

Conference_Location

Izmir

Print_ISBN

978-1-4577-1685-0

Type

conf

DOI

10.1109/CEM.2011.6047338

Filename

6047338