.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
.png)
P42 - A Python Dynamical Core for Operational Numerical Weather Prediction
Description
Numerical weather prediction is vital for applications like population warnings and energy predictions. However, adapting forecasts to diverse hardware poses challenges. MeteoSwiss relies on the ICON model up to a one km resolution, initially ported to GPUs using OpenACC. While enabling GPU use, OpenACC+Fortran has limitations in portability and maintenance.
Exploring alternatives, we focus on the EXCLAIM project, targeting the dynamical core (55% of runtime). Implementing the dynamical core's computational stencils with gt4py Python departs from Fortran traditions. Our work details this shift, emphasizing the productivity gains with this new Python framework.
We present optimizations and compare the Python-based dynamical core with the base OpenACC version, highlighting computational efficiency and development ease. Acknowledging challenges, especially in operational weather prediction.
Presenter(s)
Presenter
After completing a master's degree in computational science I briefly worked on the Open Earth Compiler, a stencil compiler for weather and climate codes, supervised by Tobias Gysi and Tobias Grosser. I now work at Meteoswiss in the context of the GLORI and EXCLAIM projects with the goal to develop the next generation weather and climate models in a higher level language - in our case Python - to be able to abstract hardware details away, leaving portability and performance to automatic code generation.