Porting an Explicit Time-Domain Volume Integral Equation Solver onto Multiple GPUs Using MPI and OpenACC
Keywords:
Explicit marching-on-in-time scheme, GPU, MPI, OpenACC, time-domain volume integral equationAbstract
A scalable parallelization algorithm to port an explicit marching-on-in-time (MOT)-based time domain volume integral equation (TDVIE) solver onto multi-GPUs is described. The algorithm makes use of MPI and OpenACC for efficient implementation. The MPI processes are responsible for synchronizing and communicating the distributed compute kernels of the MOT-TDVIE solver between the GPUs, where one MPI task is assigned to one GPU. The compiler directives of the OpenACC are responsible for the data transfer and kernels’ offloading from the CPU to the GPU and their execution on the GPU. The speedups achieved against the MPI/OpenMP code execution on multiple CPUs and parallel efficiencies are presented.
Downloads
References
CUDA, www.nvidia.com, 2016.
OpenCL, www.khronos.org/opencl, 2016.
OpenACC, www.openacc-standard.org, 2016.
OpenMP, www.openmp.org, 2016.
A. Al-Jarro, M. A. Salem, H. Bagci, T. M. Benson, P. Sewell, and A. Vukovic, “Explicit solution of the time domain volume integral equation using a predictor-corrector scheme,” IEEE Trans. Antennas Propag., vol. 60, no. 11, pp. 5203-5214, 2012.
N. T. Gres, A. A. Ergin, E. Michielssen, and B. Shanker, “Volume-integral-equation-based analysis of transient electromagnetic scattering from threedimensional inhomogeneous dielectric objects,” Radio Sci., vol. 36, no. 3, pp. 379-386, May 2001.
Y. Liu, A. Al-Jarro, H. Bagcı, and E. Michielssen, “Parallel PWTD-accelerated explicit solution of the time domain electric field volume integral equation,” IEEE Trans. Antennas Propag., vol. 64, no. 6, pp. 2378-2388, 2016.
A. Al-Jarro, M. Cheeseman, and H. Bagci, “A distributed-memory parallelization of the explicit time-domain volume integral equation solver using a rotating tiles paradigm,” in Proc. 28th Int. Review of Progress in Appl. Comp. Electromagn., 2012.
A. Al-Jarro and H. Bagci, “An unstructured mesh partitioning scheme for efficiently parallelizing an explicit time domain volume integral equation solver,” in Proc. 29th Int. Review of Progress in Appl. Comp. Electromagn., 2013.
S. Feki, A. Al-Jarro, A. Clo, and H. Bagci, “Porting an explicit time-domain volume-integral-equation solver on GPUs with OpenACC,” IEEE Antennas Propag. Mag., vol. 56, pp. 265-277, 2014.
S. Feki, A. Al-Jarro, and H. Bagci, “Multi-GPUbased acceleration of the explicit time domain volume integral equation solver using MPIOpenACC,” in Proc. IEEE Int. Symp. Antennas Propag. and USNC/URSI National Radio Sci., Meet., 2013.
