A Practical Look at GPU-Accelerated FDTD Performance
Keywords:
A Practical Look at GPU-Accelerated FDTD PerformanceAbstract
This paper outlines several key features and conditions that impact the performance of FDTD on GPUs. It includes relevant performance measurements as well as practical suggestions on how to mitigate their impact. Among these factors are: PML depth, the number of unique materials, dispersive materials, the impact of field reads/observations, simulation orientation, and domain decomposition using multiple GPUs. The paper shows that the performance of FDTD on GPUs can be limited in certain extreme cases, but with proper care on the part of the designer these cases can be managed and maximum performance guaranteed.
Downloads
References
K. S. Yee, “Numerical Solution of Initial
Boundary Value Problems Maxwell’s
Equation in Isotropic Media”, IEEE Trans.
Antennas and Prop., Vol. 14, No. 3, pp. 302-
, 1966.
S. E. Krakiwsky, L. E. Turner, M.
Okoniewski, “Acceleration of Finite-
Difference Time-Domain (FDTD) Using
Graphics Processor Units (GPU)”, IEEE MTT-
S Int. Symposium Digest, Vol. 2, pp. 1033-
, 2004.
P. F. Curt, J. P. Durbano, M. R. Bodnar, S.
Shi, M. S. Mirotznik “Enhanced Functionality
for Hardware-Based FDTD Accelerators,”
ACES Journal, Vol. 22 No. 1, 2007.
P. Sypek, M. Mrozowski, "Optimization of a
FDTD code for graphical processing units,"
The 17th. Int. Conf. on Microwaves, Radar
and Wireless Communications, MIKON, May
A. Taflove, S. Hagness. Computational
Electrodynamics: The Finite-Difference Time-
Domain Method, 3 rd ed., Artech House, 2005.
M. J. Inman, A. Z. Elsherbeni, J. G. Maloney,
and B.N. Baker, “GPU Based FDTD Solver
with CPML Boundaries,” IEEE Antennas and
Propagation Society International Symposium,
pp. 5255- 5258, 2007.
J. A. Roden and S. D. Gedney, “Convolutional
PML (CPML): An efficient FDTD
implementation of the CFS-ML for arbitrary
media”, IEEE Transactions on Antennas and
Propagation, Vol. 50, 2002, pp. 258-265.
C. Ong, M. Weldon, D. Cyca, and M.
Okoniewski, “Acceleration of large-scale
FDTD simulations on high performance GPU
clusters,” IEEE Antennas and Propagation
Society International Symposium, APSURSI
'09, pp. 1 – 4, 2009
