Jacob Page and Moritz Linkmann were part of a successful bid to the ExCALIBUR programme which aims to future-proof high priority simulation codes and algorithms

Jacob Page and Moritz Linkmann were part of a successful bid led by EPCC to the ExCALIBUR (Exascale Computing: ALgorithms and Infrastructures Benefiting UK Research) programme which aims to future-proof high priority simulation codes and algorithms to meet the rapid evolution of supercomputing technology. It is led by the UK Met Office and EPSRC along with the Public Sector Research Establishment, the UK Atomic Energy Authority (UKAEA) and UK Research and Innovation (UKRI) research councils.

Through ExCALIBUR, £600,000 has been secured for a three-year project to enable the use of Machine Learning (ML) techniques in large-scale physical simulations with the main focus on fluid dynamics and turbulent flows. Such simulations are used for example in weather and climate prediction, or for modelling magnetohydrodynamics instabilities in tokamak fusion reactors. EPCC will provide access to and experience with novel computing clusters such as ARCHER2 and Cerebras CS-1 as well as knowledge and technology transfer and software development, while the School of Mathematics will supply expertise in the fields of magnetohydrodynamics and turbulence modelling.

The resulting tool will provide an efficient and scalable C++ interface accessible from traditional simulation codes that will support researchers in applying ML techniques to relevant parts of their simulations, giving them the freedom to experiment with various ML techniques without the need to learn new programming languages and frameworks (such as Python and TensorFlow).

This work is funded as part of the UK Met Office's Strategic Priorities Fund programmes, which will bring together a broad range of research disciplines to drive an increase in high quality multi- and interdisciplinary research and innovation.