ESCAPE-2 Energy-efficient SCalable Algorithms for weather and climate Prediction at Exascale European Commission
Keywords
High Performance Computing
Advanced Numerical Methods for Scientific Computing
Geosciences/Protection of Land and Water Resources
Contact:
Luca Bonaventura
MOX responsible:
Luca Bonaventura
Team:
Tommaso Benacchio
Start date:
October 2018
End date:
September 2021
Scientific documentation:
http://www.hpc-escape2.eu
Abstract:
ESCAPE-2 will develop world-class, extreme-scale computing capabilities for European operational numerical weather and climate prediction systems. It continues the pioneering work of the ESCAPE project. The project aims to attack all three sources of enhanced computational performance at once, namely (i) developing and testing bespoke numerical methods that optimally trade off accuracy, resilience and performance, (ii) developing generic programming approaches that ensure code portability and performance portability, (iii) testing performance on HPC platforms offering different processor technologies. ESCAPE-2 will prepare weather and climate domain benchmarks that will allow a much more realistic assessment of application specific performance on large HPC systems than current generic benchmarks such as HPL and HPCG. These benchmarks are specifically geared towards the pre-exascale and exascale HPC infrastructures that the European Commission and Member States will invest in through the European Commission Joint Undertaking.
The role of MOX in ESCAPE-2 is the development of efficient semi-implicit solvers for Discontinuous Galerkin discretizations of the Euler equations. ESCAPE-2 is funded by the European Commission under the Future and Emerging Technologies - High-Performance Computing call for research and innovation actions, grant agreement 800897.