Comparison of some parallelisation strategies of thermalhydraulics codes on GPUs
CEA-Saclay, DEN/DANS/DM2S/STMF, Commissariat a l’Energie atomique et aux Energies alternatives, F-91191 Gif-sur-Yvette, France
Modern supercomputers architecture is now often based on hybrid concepts combining parallelism to distributed memory, parallelism to shared memory and also to GPUs. In this work, we propose a new approach to take advantage of these graphic cards in TH algorithms.
Key words: GPU / unstructured grid / multithreading / CUDA / OpenCL
© Owned by the authors, published by EDP Sciences, 2014