A. V. Gorobets, S. A. Soukov, P. B. Bogdanov, “Towards hybrid supercomputing in computational fluid dynamics”, Informatsionnye Tekhnologii i Vychslitel'nye Sistemy, 2013, Issue 4,Pages <nobr>15

HIGH-EFFICIENCY COMPUTING SYSTEMS

Towards hybrid supercomputing in computational fluid dynamics

A. V. Gorobets^a, S. A. Soukov^a, P. B. Bogdanov^b

^a Keldysh Institute of Applied Mathematics of Russian Academy of Sciences, Moscow
^b Scientific Research Institute for System Analysis of the Russian Academy of Sciences, Moscow

Abstract: The work is devoted to the efficient use of modern hybrid supercomputers based on massively-parallel computing accelerators used as math co-processors. Adaptation of algorithms is demonstrated on example of computational fluid dynamics finite-volume methods for unstructured meshes. Different kinds of devices are considered including GPU of NVIDIA and AMD, Intel Xeon Phi accelerators.

Keywords: CFD, unstructured mesh, MPI, OpenMP, OpenCL, GPU.