Adaptive Mesh Refinement GPU Implementation using Dynamic Memory Arrangement for Interface Advection by Conservative Phase Field Method

摘要

Adaptive Mesh Refinement (AMR) is a widely used method in computational ruid dynamics (CFD) to save thernmemory usage and reduce the number of elements of the computational domain, in order to enlarge the scale ofrnthe domain as well as increase the performance. The highly parallelism computation power of graphics processingrnunits (GPU) have been widely utilized in accelerating the computations. However, due to the diu000eculties and thernbig overheads of dynamic changes of the AMR blocks during the simulation, an eu000ecient AMR implementationrnusing GPU is a challenging task. In this paper, we present an eu000ecient GPU implementation of tree-based AMRrnusing dynamic memory arrangement and apply it to an interface advection problem, solving by a conservativernphase feld method. Performance measurement showed an obviously 16.87u0002 speedup of our GPU implementationrncompares to a OpenMP CPU implementation.