2D lid-driven cavity flow simulation using GPU-CUDA with a high-order finite difference scheme

Abstract

The high parallelism and low cost of the graphic processing units (GPUs) have attracted the interest of scientists and engineers requiring high computational power with a modest investment. This work explores the use of a GPU in the solution of the 2D lid-driven cavity flow problem using the pressure–velocity formulation for Reynolds numbers up to 10,000 and turning to a 4th order finite difference scheme for spatial discretization. Results showed good agreement with those reported in the literature. The solver was implemented in both the CPU and the GPU in order to compare their performance, whereupon the latter was seventy times faster