Complete PISO and SIMPLE solvers on Graphics Processing Units

TL;DR

This paper presents GPU implementations of PISO and SIMPLE solvers for Navier-Stokes equations, introducing a new sparse matrix format and demonstrating significant speedups over CPU-based methods.

Contribution

The authors developed GPU-optimized PISO and SIMPLE solvers with a novel sparse matrix format, achieving substantial computational speedups for CFD simulations.

Findings

GPU implementation outperforms CPU by a factor of 4.2

New sparse matrix format enhances CFD operation efficiency

Validated on standard steady and unsteady problems

Abstract

We implemented the pressure-implicit with splitting of operators (PISO) and semi-implicit method for pressure-linked equations (SIMPLE) solvers of the Navier-Stokes equations on Fermi-class graphics processing units (GPUs) using the CUDA technology. We also introduced a new format of sparse matrices optimized for performing elementary CFD operations, like gradient or divergence discretization, on GPUs. We verified the validity of the implementation on several standard, steady and unsteady problems. Computational effciency of the GPU implementation was examined by comparing its double precision run times with those of essentially the same algorithms implemented in OpenFOAM. The results show that a GPU (Tesla C2070) can outperform a server-class 6-core, 12-thread CPU (Intel Xeon X5670) by a factor of 4.2.