Heterogeneous Computing on Mixed Unstructured Grids with PyFR

TL;DR

PyFR is an open-source solver capable of high-order accurate unsteady flow simulations on mixed unstructured grids across heterogeneous multi-node hardware, demonstrating good performance and accuracy.

Contribution

This paper demonstrates the ability of PyFR to perform high-order unsteady flow simulations on mixed unstructured grids using heterogeneous multi-node hardware.

Findings

PyFR successfully simulated flow over a circular cylinder at Re=3900.

Performance and accuracy were validated on CPU and GPU hardware.

PyFR is freely available under an open-source license.

Abstract

PyFR is an open-source high-order accurate computational fluid dynamics solver for mixed unstructured grids that can target a range of hardware platforms from a single codebase. In this paper we demonstrate the ability of PyFR to perform high-order accurate unsteady simulations of flow on mixed unstructured grids using heterogeneous multi-node hardware. Specifically, after benchmarking single-node performance for various platforms, PyFR v0.2.2 is used to undertake simulations of unsteady flow over a circular cylinder at Reynolds number 3 900 using a mixed unstructured grid of prismatic and tetrahedral elements on a desktop workstation containing an Intel Xeon E5-2697 v2 CPU, an NVIDIA Tesla K40c GPU, and an AMD FirePro W9100 GPU. Both the performance and accuracy of PyFR are assessed. PyFR v0.2.2 is freely available under a 3-Clause New Style BSD license (see www.pyfr.org).