06.05.2022

ESiWACE scientists take part in the 7th HPC workshop of the European Network for Earth System modelling (ENES), May 9th -11th, 2022 in Barcelona (ES). It is organised by IS-ENES3 in collaboration with ESiWACE and will gather experts on HPC in earth system modelling. Here are some ESiWACE related talks:

Session 1: European and International HPC Landscape

  • In "ARMing the IFS: Experiments and experiences from porting the ECMWF model to Fugaku" Sam Hatfield (ECMWF) will provide an overview of his latest results, including an honest discussion of difficulties encountered but also the benefits of the Fujitsu A64FX ARM CPU powering the Fugaku supercomputer, currently the fastest machine in the world.

Session 2: Mix traditional modeling with ML

  • "Machine learning for weather and climate predictions" by Peter Dueben (ECMWF) will give a summary of the latest developments in machine learning for weather and climate predictions as discussed in the recent workshop on machine learning at the European Centre for Medium-Range Weather Forecasts (ECMWF). The talk will outline how machine learning, and in particular deep learning, could help to improve weather predictions in the coming years, name challenges for the use of machine learning, and suggest developments (research/software/hardware) that should enable the community of Earth system modelling to make quick progress.

Session 3: Performance

  • "Performance optimization in NEMO ocean model", Italo Epicoco (CMCC) shows the most recent developments in one of Europe's flagship models.

  • "The new load balancing tool in OASIS3-MCT_5.0" by Eric Maisonnave (CERFACS), will introduce the new features of the OASIS-MTC coulpler and will discuss its usefulness in the age of Exascale.

  • "The Potential of Functional Concurrency in W&C models" by R. Budich (MPI-M) will give an introduction to FC and report on its application.

Session 4: Heterogeneous architectures (accelerators)

  • "Preparing ICON for heterogeneous architectures - Experiences and the way forward" by Claudia Frauen (DKRZ) will give insights into the most recent developments on one of the ESiWACE flagship models.

  • In "PSyclone in Met Office: Evolution and revolution" Iva Kavcic (Met Office) will give an overview on how the domain-specific compiler and source-to-source translation "PSyclone" is applied to various use cases.

  • "Accelerating tracer transport in FESOM-2 with GPU's", G. van den Oord (NLeSC) will give insights into the ESiWACE Service 1 on porting applications to pre-exascale machines granted for the FESOM-2 team

Session 5: Data Worflows

  • "Towards HPC and Big Data convergence for climate analysis at scale" by Donatello Elia (CMCC) will introduce some relevant challenges towards large-scale HPDA for eScience and describe some of the developments for the Ophidia framework to address these challenges.

For additional information and the agenda, please go to the official event page on the ENES website.

Although registration is closed, a few virtual places are left. If you are interested, please contact the organiser sophie.morellon[a]lsce.ipsl.fr.