GPU Acceleration of the Advanced Regional Prediction System (ARPS)

A recent breakthrough in high-performance computing is the application of general-purpose graphics processing units (GPGPUs) to scientific computing. This paper presents a case study of a project to reformulate mathematical com-putations within a weather model, the Advanced Regional Prediction System (ARPS), to work with GPGPU hard-ware. This hardware typically consists of hundreds of simple processors, compared to the conventional central processing units (CPUs) with fewer than eight processors (as found in most computers). While GPGPUs are extremely power-ful, their usage requires specialized programming to achieve their full potential. As a forecasting tool, ARPS is capa-ble of producing very high-resolution weather simulations. Using ARPS to simulate detailed atmospheric disturbances necessitates the use of large-scale distributed-memory paral-lel computing clusters. The adaptation of a critical numer-ical kernel within ARPS for GPGPUs resulted in a six-fold speedup over the CPU version. These optimizations dra-matically reduce simulation time, thereby leading to faster weather predictions that may benefit society as well as the research community