Analysis of FastEddy Model Data on GPUs
TimeThursday, 19 November 20208:30am - 5pm EST
DescriptionData analysis of atmospheric model outputs is often embarrassingly parallel and compute intensive, and is traditionally done on central processing units (CPUs). FastEddy is a General Purpose Graphical Processing Units (GPGPU) -based Large Eddy Simulation (LES) atmospheric model developed at NCAR-Research Application Laboratory that solves fundamental dynamical equations and computes turbulence at a high resolution, producing large datasets that reside on GPUs. To reduce the amount of data movement from GPUs to CPUs for analysis, and to reduce the need for end-users to learn complex GPU programming using CUDA, this work explores performing the analysis on GPUs using only Python libraries.