Operational numerical weather prediction with ICON on GPUs (version 2024.10)
Abstract. Numerical weather prediction and climate models require continuous adaptation to take advantage of advances in high-performance computing hardware. This paper presents the port of the ICON model to GPUs using OpenACC compiler directives for numerical weather prediction applications. In the context of an end-to-end operational forecast application, we adopted a full-port strategy: the entire workflow, from physical parameterizations to data assimilation, was analyzed and ported to GPUs as needed. Performance tuning and mixed-precision optimization yield a 5.6x speed-up compared to the CPU baseline in a socket-to-socket comparison. The ported ICON model meets strict requirements for time-to-solution and meteorological quality, in order for MeteoSwiss to be the first national weather service to run ICON operationally on GPUs with its ICON-CH1-EPS and ICON-CH2-EPS ensemble forecasting systems. We discuss key performance strategies, operational challenges, and the broader implications of transitioning community models to GPU-based platforms.
Authors
- Daniel Leuenberger (ORCID: https://orcid.org/0009-0002-8785-1010)
- Oliver Fuhrer (ORCID: https://orcid.org/0000-0002-0682-1374)
- P. Kaufmann (ORCID: https://orcid.org/0000-0002-0873-598X)
- Michael Jähn (ORCID: https://orcid.org/0000-0003-0886-8245)
- Marek Jacob (ORCID: https://orcid.org/0000-0001-9059-6458)
- Fabian Gessler
- Andreas Pauling
- Claire Merker
- David Leutwyler (ORCID: https://orcid.org/0000-0002-5141-1737)
- William H. Sawyer (ORCID: https://orcid.org/0000-0001-5894-0765)
- Sergey S. Kosukhin
- Ulrich Schättler
- Lukas Jansing (ORCID: https://orcid.org/0000-0003-4644-5403)
- Annika Lauber (ORCID: https://orcid.org/0000-0001-5731-2652)
- Xavier Lapillonne (ORCID: https://orcid.org/0000-0001-6114-4321)
- Ralf Müller (ORCID: https://orcid.org/0000-0002-6251-6681)
- T. C. Schulthess (ORCID: https://orcid.org/0000-0002-1086-5812)
- Victoria Cherkas
- Jonas Jucker
- Peter Baumann (ORCID: https://orcid.org/0000-0003-3860-4726)
- Benjamin Cumming (ORCID: https://orcid.org/0009-0004-1524-2560)
- Dmitry Alexeev (ORCID: https://orcid.org/0000-0003-0783-1176)
- Guillaume Van Parys (ORCID: https://orcid.org/0000-0001-8799-9555)
- Carlos Osuna (ORCID: https://orcid.org/0000-0002-9896-9957)
- Daniel Hupp
- Gabriel Vollenweider
- P. Steiner
- Remo Dietlicher (ORCID: https://orcid.org/0000-0003-0217-7232)
- Christoph Müller (ORCID: https://orcid.org/0000-0002-9491-3550)
- André Walser
- M. Arpagaus (ORCID: https://orcid.org/0000-0002-4159-4031)
- Marco Induni
- Matthias Kraushaar
- Mikael Stellio
Institutions
- Federal Office of Meteorology and Climatology MeteoSwiss (CH)
- Deutscher Wetterdienst (DE)
- German Climate Computing Centre (DE)
- ETH Zurich (CH)
- Short and Associates (United States) (US)
- Max Planck Institute for Meteorology (DE)
- Swisscom (Switzerland) (CH)
Publication Details
- Published
- 2025-08-25
- DOI
- https://doi.org/10.5194/egusphere-2025-3585
- Citations
- 2
- Primary Topic
- Precipitation Measurement and Analysis
- Type
- article
- Field-Weighted Citation Impact
- 2.60
Funders
- Deutsches Klimarechenzentrum