New Exascale collaborations at EPCC
ExCALIBUR is a £45.7m programme to address the challenges and opportunities offered by computing at the exascale (that is, high performance computing at 1018 floating point operations per second). EPCC is a partner in three of the programme's eight groups.
ExCALIBUR is led by the Met Office and the Engineering and Physical Sciences Research Council, and is a UK Research and Innovation (UKRI) Strategic Priorities Fund programme. The Strategic Priorities Fund is one of the UK Government’s largest programmes to work on multi-disciplinary and inter-disciplinary research and innovation, and has the ultimate aim of boosting the productivity and competitiveness of the economy. In the context of ExCALIBUR, this includes work on currently intractable problems of strategic importance such as drug and vaccine research, climate and weather prediction, and fusion power and green energy sources.
Specifically, the aim is the redesign and reimplementation of high priority codes and algorithms to maintain a leading position for the UK in high performance computing.
With the first exascale machines perhaps becoming available between 2021–2023, the challenges are certainly formidable. A range of development is required from systems software and libraries, applications themselves, and utilities to prepare, analyse and visualise large amounts of data.
A very high degree of parallelism will be relevant throughout – perhaps O(109)-O(1010) simultaneous threads of execution will be employed at the exascale. Applications must maximise efficient use of computational resources, minimise costly data transfers and synchronisations, and at the same time deal with significant algorithmic complexity.
All this must be done with an eye not just to flexible and efficient computation and robust numerical results, but to productivity of those doing the development and to energy efficiency at the time of execution.
In October 2019, as the first stage in ExCALIBUR, UKRI put out a call for high priority use cases around which to form design and development working groups. EPCC is now active, in collaboration with universities and industrial partners around the UK, in three of the resultant eight groups which have been funded (see below).
The three groups will concentrate on different areas: mesh generation and manipulation for finite element problems, materials simulations, and lattice field theory. It will be the results of such programmes which will enable the benefits of efficient exascale computing to be achieved.
EPCC’s ExCALIBUR collaborations
The Materials and Molecular Modelling Exascale Design and Development Working Group
This project brings together Research Software Engineers and experts from mathematics and computer science with domain experts in Materials and Molecular Modelling (MMM) to ensure this key research area is ready to exploit future exascale computing resources. This will cover all components required by the community as pre-exascale and exascale computational resources become available to researchers. The project will equip the UK MMM community with the tools and communities required to use exascale computational resources efficiently to address many EPSRC Grand Challenges.
The project is led by UCL with EPCC among the co-leaders.
ELEMENT: the Exascale Mesh Network
ELEMENT addresses the high priority use case of meshing for the Exascale (ensuring that meshes are of sufficient quality to represent Exascale problems and can be partitioned efficiently to minimise load imbalance) as well as meshing at the Exascale (creating highly scalable solutions able to exploit extreme levels of parallelism).
Meshing and geometry management remain a significant bottleneck for complex applications on HPC platforms, posing a challenging obstacle that must be overcome to enable Exascale simulations. From a technical perspective, these issues include improved geometric handling, mesh adaptation and optimisation, intelligent meshing, automation and robustness, all within a large distributed environment that lies outside of our current capabilities.
The project is led by EPCC.
EXALAT: Lattice Field Theory at the Exascale Frontier
EXALAT is a collaboration led by the University of Edinburgh, with total funding of £365k to build on the recognised expertise in the Lattice Field Theory (LFT) community to use High-end Computing, to inform and upskill the wider scientific ecosystem for Exascale computing.
The project will develop a roadmap for LFT, as an example for other domains, to prepare for science on the first Exascale computers. It will also develop initial resources (best practice, training, and algorithmic/ software templates) to begin the journey.