Thu13Jun201916:45 - 18:45Zurich, Switzerland
PASC 2019 - Modeling Cloud Physics: Preparing for Exascale
Accelerating Cloud Physics and Atmospheric Models using GPUs, KNLs and FPGAs
Nick Brown - EPCC, 17:30 - 18:00
The Met Office NERC Cloud model (MONC) is an atmospheric model used to study clouds and turbulent flows. This is coupled with the CASIM microphysics model which provides the capability to investigate interactions at the millimetre scale and study the formation and development of moisture. Crucially, MONC and CASIM are both heavily computationally intensive, and in this talk I will explore the role that accelerators might provide for improving the capability of these code. Using OpenACC for GPU acceleration, the approach adopted for MONC, where we were able to isolate specific kernels, differed from the approach for CASIM where we were forced to offload the entirety of this microphysics model due to the tightly coupled nature of the code. Whilst there were several caveats and lessons learnt, experiments on Piz Daint demonstrated a benefit to using GPUs for these models. We will also consider the role other HPC technologies, namely KNL and FPGAs, might have to play in accelerating cloud physics codes at exascale. For the later, as part of the EXCELLERAT CoE, we have ported aspects of the MONC model onto FPGAs and will detail the challenges, along with performance characteristics, contrasted against using CPUs, GPUs and KNL.