Development of efficient GPU parallelization of WRF Yonsei University planetary boundary layer scheme

Abstract

Abstract.The planetary boundary layer (PBL) is the lowest part of the atmosphere and where its character is directly affected by its contact with the underlying planetary surface. The PBL is responsible for vertical sub-grid-scale fluxes due to eddy transport in the whole atmospheric column. It determines the flux profiles within the well-mixed boundary layer and the more stable layer above. It thus provides an evolutionary model of atmospheric temperature, moisture (including clouds), and horizontal momentum in the entire atmospheric column. For such purposes, several PBL models have been proposed and employed in the weather research and forecasting (WRF) model of which the Yonsei University (YSU) scheme is one. To expedite weather research and prediction, we have put tremendous effort into developing an accelerated implementation of the entire WRF model using graphics processing unit (GPU) massive parallel computing architecture whilst maintaining its accuracy as compared to its central processing unit (CPU)-based implementation. This paper presents our efficient GPU-based design on a WRF YSU PBL scheme. Using one NVIDIA Tesla K40 GPU, the GPU-based YSU PBL scheme achieves a speedup of 193× with respect to its CPU counterpart running on one CPU core, whereas the speedup for one CPU socket (4 cores) with respect to 1 CPU core is only 3.5×. We can even boost the speedup to 360× with respect to 1 CPU core as two K40 GPUs are applied.

Weather forecast using WRFg at 2 km resolution over two provinces in northwestern Panama

“…Weather forecasts are usually run on complex arrays of supercomputers called high-performance computing (HPC) systems. In this sense, in the field of meteorology, the computational power of graphics processing units (GPUs) has been incorporated into HPC systems, obtaining considerable reductions in the execution time of simulations [Michalakes et al (2008), Sharma et al (2015) and Huang et al (2015)]. These types of technologies allow lower initial costs, both of the physical facilities and of the computational infrastructure of a laboratory, since this new equipment have smaller physical dimensions and a higher computational capacity.…”

Implementation of a scalable, performance portable shallow water equation solver using radial basis function-generated finite difference methods

“…Most are implemented using message passing interface (MPI) and OpenMP and optimized strictly for standard CPU systems, resulting in either bad performance or complete incompatibility on accelerated architectures. Some atmospheric models have been partially converted to compute unified device architecture (CUDA) for compatibility with NVIDIA GPUs (Huang et al, 2015), while others have adopted a strategy of creating locally developed domain specific tools and frameworks (Fuhrer et al, 2014). Both approaches appear to require significant software development and maintenance investments, as well as the risk of a good deal of duplication of effort between centers with different models.…”