Title :
Low-Frequency MLFMA on Graphics Processors
Author :
Cwikla, M. ; Aronsson, J. ; Okhmatovski, V.
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. of Manitoba, Winnipeg, MB, Canada
fDate :
7/2/1905 12:00:00 AM
Abstract :
A parallelization of the low-frequency multilevel fast multipole algorithm (MLFMA) for graphics processing units (GPUs) is presented. The implementation exhibits speedups between 10 and 30 compared to a serial CPU implementation of the algorithm. The error of the MLFMA on the GPU is controllable down to machine precision. Under the typical method-of-moments (MoM) error requirement of three correct digits, modern GPUs are shown to handle problems with up to 7.5 million degrees of freedom in dense matrix approximation.
Keywords :
computer graphic equipment; coprocessors; method of moments; multivalued logic; GPU; dense matrix approximation; graphics processing units; graphics processors; low-frequency MLFMA; low-frequency multilevel fast multipole algorithm; machine precision; method-of-moments error; serial CPU implementation; CUDA; fast algorithms; graphics processing unit (GPU); low-frequency fast multipole method; multiscale modeling;
Journal_Title :
Antennas and Wireless Propagation Letters, IEEE
DOI :
10.1109/LAWP.2010.2040571
