GPU Acceleration of Algebraic Multigrid for Low-Frequency Finite Element Methods
Thayer School Of Engineering
Dartmouth College, 8000 Cummings Hall, Hanover, NH 03755, USA
Dartmouth College, 8000 Cummings Hall, Hanover, NH 03755, USA
IEEE Antennas and Propagation Society International Symposium (APSURSI), 2012
@inproceedings{,
year={2012},
authors={Attardo, Elia A and Borsic, Andrea},
title={GPU acceleration of Algebraic Multigrid for low-frequency finite element methods},
booktitle={Antennas and Propagation Society International Symposium (APSURSI), 2012 IEEE},
pages={1–2},
organization={IEEE},
number={attardo2012gpu}
}
This paper introduces a GPU acceleration of a Wavelet-based Algebraic Multigrid used as preconditioner for solving the Laplace’s equation discretized by Finite Element Method. We conduct some tests using a CPU-based direct solver, a CPU-based Preconditined Conjugate Gradient (PCG), and a GPU-based PCG. Finally, we report the solution time and the speed-up achieved in solving the discretized problem.
June 9, 2013 by egpu
