GPU-Based Acceleration for FDTD Method of Acoustic Field Analysis

Abstract:

The finite difference time domain (FDTD) method is a numerical technique used in time-domain acoustic simulation. However, the huge computing capacity has been an important limiting factor for its applications. In this paper, we will present a graphics processing units (GPU) - based parallel FDTD framework and its successful application to the estimation of time-series sound pressure data. Compared with the computer, the runtime of acoustic field FDTD simulation on GPU is dramatically decreased. The acceleration in runtime has made such study possible, and will pave the way for other studies of large-scale computational acoustic problems which were previously impractical.

• Periodical: Applied Mechanics and Materials (Volumes 602-605)
• Pages: 3702-3705
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.602-605.3702
• Online since: August 2014
• Authors: Ze Rui Yang*, Jie Yang
• Keywords: Acoustic Simulation, FDTD, GPU

References

[1] K. S. Yee in: IEEE Trans. Antennas and Propagation 14 (1996), pp.302-307.

[2] A. Taflove, Computational Electrodynamics: The Finite-Difference Time-Domain Method, Artech House, Boston, (1995).

[3] Naoki Kawada, Kan Okubo, Norio Tagawa and Takao Tsuchiya in: Proceedings of 20th international congress on acoustic. (2010), pp.1-6.

[4] D. Botteldooren in: J. Acoust. Soc. Am. 100(1996), pp.1357-1367.

[5] I.M. Hallaj and R.O. Cleveland in: The Journal of the Acoustical Society of America, 105(1999), pp.7-12.

[6] Information on http: /www. nvidia. com/object/tesla-workstations. html.

[7] Q. -H. Liu and J. Tao in: J. Acoust. Soc. Am. 102(1997), p.2072-(2082).