Benchmark suite for PyFR

The objective here is to develop a benchmark for multi node multi-gpu runs for PyFR [1].

The benchmarks are based on the supplementary material of [2]. Minor changes to the files had to be introduced to keep up with the PyFR version. The target architecture is Power8 with 4 GPUs each. There are some specific LSF Data Mover commands due to the adoption of DMD at Daresbury.

Isentropic vortex

WIP

Cylinder 3D

SD7003

WIP

References

1. PyFR: An Open Source Framework for Solving Advection-Diffusion Type Problems on Streaming Architectures using the Flux Reconstruction Approach. F. D. Witherden, A. M. Farrington, P. E. Vincent. Computer Physics Communications, Volume 185, Pages 3028-3040, 2014.

2. B.C. Vermeire, F.D. Witherden, P.E. Vincent, On the utility of GPU accelerated high-order methods for unsteady flow simulations: A comparison with industry-standard tools, Journal of Computational Physics, Volume 334, 1 April 2017, Pages 497-521, ISSN 0021-9991, http://dx.doi.org/10.1016/j.jcp.2016.12.049.