Research Article
Parallelization of an Unsteady ALE Solver with Deforming Mesh Using OpenACC
Table 4
Performance comparisons of multicore CPU and GPU results on mesh deformation algorithm.
| | 2D case | | Single CPU core | GPU | 20 OpenMP threads |
| | 5672268 nodes | Matrix assembly | 9.8 s | 0.23 s | 0.7 s | | 11117646 triangles | P-BiCGSTAB | 52 s | 1.63 s | 4.19 s |
| | 3D case | | Single CPU core | GPU | 20 OpenMP threads |
| | 4127686 nodes | Matrix assembly | 12.6 s | 0.33 s | 0.78 s | | 24023132 tetrahedrons | P-BiCGSTAB | 58.1 s | 1.84 s | 4.83 s |
|
|
