Webinar: Using GPUs to accelerate National Weather Forecasts - Challenges and state-of-the-art practices

Name: Webinar: Using GPUs to accelerate National Weather Forecasts - Challenges and state-of-the-art practices
Start: 2024-06-27T12:30:00+01:00
End: 2024-06-27T14:00:00+01:00
Location: Online

Programme

ICON's Operational GPU Integration by Xavier Lapillonne, MeteoSwiss: Explore the rewriting of NWP models for GPU efficiency, illustrated by the recently operational ICON project at MeteoSwiss. Learn about the integration of ICON into the new ALPS High-Performance Computing Platform at CSCS, addressing operationalization and maintenance challenges.
Code Modularization and Modernization by Xavier Lapillonne, MeteoSwiss: Explore the ICON-C project's refined approach, which introduces a clear separation of responsibilities for different code parts, enabling faster and more flexible development while optimizing modules for efficiency.
Performance and Adaptability with Domain Specific Language (DSLs) by Christoph Müller, MeteoSwiss: Witness the shift to descriptive Python-based user codes, empowered by DSLs like GT4Py, for achieving performance portability and adaptability. Continuing our focus on redesigning and refactoring model codes, this effort emphasizes creating architecture-agnostic user codes through code splitting. We'll concentrate on addressing the challenges of maintaining DSLs for operational implementation.
Starting anew with the Development of Momentum® Weather and Climate Model by Iva Kavcic, Met Office: We explore new horizons with the Next Generation Modelling System Programme from the Met Office, paving the way for a new dynamical core and software infrastructure in weather and climate modelling. Learn firsthand about their experience using DSLs and automatic code parallelism to enhance operational performance, with plans for implementation from the mid-2020s onwards.
NWP Models on AMD GPUs by Bentorey Hernandez Cruz, ECMWF: Learn about the ongoing development of NWP models on AMD GPUs, exemplified by the DestinE project implementing IFS on the LUMI supercomputer. Examine the hurdles and progress in harnessing AMD's architecture for weather and climate modeling.