Coalesced computations of the incompressible Navier–Stokes equations over an airfoil using graphics processing units

Recent progress and challenges in exploiting graphics processors in computational fluid dynamics

“…Systems of linear equations present in both stages were solved using Jacobi iteration, which, unlike the GS method, does not require special treatment for parallelization, although convergence may occur more slowly. As with other approaches using structured grids, Iman Gohari et al [47] exploited global memory coalescing to improve performance, and focused on optimizing memory performance in detail. Using single-precision calculations, they demonstrated an overall speedup (grid construction + flow solution) of up to around 40, comparing a single GPU and single CPU core.…”

“…Iman Gohari et al [47] developed a solver for the incompressible, turbulent Navier-Stokes equations, where both the grid generation stage and flow solver were performed on the GPU, and used it to simulate flow over airfoils. Using the stream function-vorticity formulation, they modeled turbulence using the Balwin-Lomax closure method.…”

Implementation and Optimization of a CFD Solver Using Overlapped Meshes on Multiple MIC Coprocessors

“…Working as coprocessors, GPUs also have been popular in CFD. Many researchers [13][14][15][16][17] have studied GPU computing on structured meshes, which involved coalesced computation technique [13], heterogeneous algorithm [15,17], numerical methods [16], etc. Corrigan et al [18] investigated an Euler solver on GPU by employing unstructured grid and gained important factor of speedup over CPUs.…”