Joint Lab ExaESM Activities
The JL-ExaESM develops fundamentally new concepts to allow necessary improvements in Earth system modeling and take advantage of future HPC architectures. It enables Exascale simulations and develops data handling concepts in order to achieve breakthroughs in the simulation of regional climate change, extreme events and their impacts on society and ecosystems. It follows the approach of co-design between computer scientists and domain scientists to address the scientific and methodological challenges of Earth system science (ESS) using simulation and data analytics methods which exploit the Exascale capabilities of future supercomputers. The JL-ExaESM provides a much-needed platform that bridges the RFs Information, Earth and Environment, and Aeronautics, Space and Transport and allows scientists from different ESS domains to develop scientifiand technological concepts for future-generation Earth system models and data analytics systems.
Following the successful exploratory phase of the Helmholtz Pilot Lab ExaESM project from October 2019 to September 2022, the initial kick-off the JL-ExaESM took place in March 2022. At the Kickoff meeting, two major scientific activities governing the future work of the lab were established. Activity #1 on Exascale code scalability seeks to enable the Exascale-readiness of ESM codes of the HGF partners, in particular by more/better parallelisation of the codes, performance portability and future-proofing, co-design and community coordination, and novel numerical methods. Activity #2 on Exascale workflow scalability focuses on the promotion of data-intensive sciences, generation of knowledge from data and models (machine learning), interactive large-scale analysis workflows, hierarchical data architectures for data-intensive simulations, and FAIR data, metadata, and data portals.
The JL-ExaESM is a “think tank” connecting computational and domain scientists to exchange at the interface of ESM- and HPC-related topics. Governed by its two new major activities, the lab aims to establish focused teams between its partners addressing specific questions and challenges of ESM at the transition to Exascale. Based on its activities, the JL-ExaESM helps building partnerships to apply for third-party funding. Considering the high degree of competition on the job market, in particular on the field of IT, the JL-ExaESM activities help try to attract the best candidates for jobs. Finally, the JL-ExaESM contributes to community building via education and user training.