GPU acceleration of linear systems for computational electromagnetic simulations

This paper has shown that simple algorithms used in many linear systems solvers can be effectively utilized on the GPU. For the case of matrix-vector multiplication it has been shown that speedups of almost 13 times can be achieved by using simple GPU methods over the speed of current CPU's. By understanding how the GPU works and creating the GPU programs to utilize its strengths, large speedups can be gained in a variety of linear systems solvers.

[1]  Pat Hanrahan,et al.  Understanding the efficiency of GPU algorithms for matrix-matrix multiplication , 2004, Graphics Hardware.

[2]  M.J. Inman,et al.  FDTD calculations using graphical processing units , 2005, IEEE/ACES International Conference on Wireless Communications and Applied Computational Electromagnetics, 2005..

[3]  John D. Owens,et al.  General Purpose Computation on Graphics Hardware , 2005, IEEE Visualization.

[4]  Jens H. Krüger,et al.  GPGPU: general purpose computation on graphics hardware , 2004, SIGGRAPH '04.

[5]  M.J. Inman,et al.  Programming video cards for computational electromagnetics applications , 2005, IEEE Antennas and Propagation Magazine.