Shared-memory, distributed-memory, and mixed-mode parallelisation of a CFD simulation code

Abstract

This paper presents some different approaches to the parallelisation of a harmonic balance Navier-Stokes solver for unsteady aerodynamics. Such simulation codes can require very large amounts of computational resource for realistic simulations, and therefore can benefit significantly from parallelisation. The simulation code addressed in this paper can undertake different modes of aerodynamic simulation and includes both harmonic balance and time domain solvers. These different modes have performance characteristics which can affect any potential parallelisation, as can the specifics of the problem being simulated. Therefore, three different techniques have been used for the parallelisation, shared-memory, distributed-memory, and a combination of the two-a hybrid or mixed-mode parallelisation. These different techniques attempt to address the different performance requirements associated with the types of simulation the code can be used for and provide the level of computational resources required for significant simulation problems. We discuss the different parallelisations and the performance they exhibit on a range of computational resources.

Evaluating the Arm Ecosystem for High Performance Computing

“…integration. It is implemented in Fortran [13] and has been parallelised using MPI [12], with each MPI process working on a set of grid blocks (geometric partitions) of the simulation.…”

Load balance and parallel I/O: Optimising COSA for large simulations

“…Scaling to large numbers of cores requires all parts of an application to be efficiently parallelised, and work to be evenly decomposed across workers. This paper outlines work undertaken [5] to take a parallelised HB CFD application, COSA [6], and enable it to scale efficiently to large numbers of cores. COSA is a finite volume NS code featuring a steady solver, a TD solver for the solution of general unsteady flows [7] [8], and a HB solver for the rapid solution of highly nonlinear unsteady periodic flows [9] [10] [11].…”

Comparative turbulent three-dimensional Navier–Stokes hydrodynamic analysis and performance assessment of oscillating wings for renewable energy applications

“…The perfect gas equation is used to link internal energy, pressure and density. The The only nonzero entries of the source term S are those of the k and ω equations, given respectively by: COSA is second order accurate in time and space, and uses a very efficient MPI parallelization [18]. The accuracy of the space-and time-discretization has been thoroughly validated by considering a wide set of analytical and experi-220 mental test cases [8,13,17].…”