GHOST master

GHOST (the Geophysical High-Order Suite for Turbulence) is an accurate and highly scalable pseudospectral code that solves a variety of PDEs often encountered in studies of turbulent flows. It is mainly developed by Pablo Mininni and Duane Rosenberg, with contributions from other users. The code uses a hybrid parallelization method combining MPI and OpenMP (GHOST also has support for GPUs using CUDA/OpenMP 5 and Intel MKL). This allows the code to run efficiently in laptops and small clusters, as well as to scale up to over 100,000 CPU cores in production runs in supercomputers. The hybrid parallelization method used in GHOST was recognized by two awards: an NCAR/CISL award, and the best paper award (technology track) at the TeraGrid 2010 conference.