Benchmarking of a 3D, unstructured, finite volume code for incompressible Navier-Stokes equation on a cluster of distributed-memory computers

The parallelization and performance of an implicit, unstructured, time-dependent computational fluid dynamics (CFD) code is described. Parallelization of the code is done within the PETSc framework using a single-program-multiple-data (SPMD) message passing model. The parallel code is shown to scale linearly within the limit of the available number of processors. A dynamic convergence criteria (DCC) is developed to enhance the performance of the code by linking the nonlinear convergence of the code and the solver convergence. The implementation of an appropriate choice of DCC in the code was shown good improvements with the speedups of over four times in terms of wall-clock time.