The UK government is investing £45 million on a new AI supercomputer created to aid scientists in modelling nuclear fusion.
The new system is set to operate at the UK Atomic Energy Authority’s (EUKAEA) Culham campus this year.
The 1.4MW system will mark the first piece of major infrastructure set out for the Culham AI Growth Zone in Oxfordshire, and is set to open in June.
Funded by the Department for Energy Security and Net Zero (DESNZ), the ‘Sunrise’ machine is set to power advances in fusion energy research using AI.
The AI is meant to enhance the ability of fusion researchers to create more detailed simulations and digital twins of complicated fusion systems, allowing for the modelling of these systems with AI prior to undertaking physical experiments which can be costly and time consuming.
The UK government says that the AI modeling capacity of the supercomputer is more advanced compared to current fusion research programmes, saying that the system is capable of 6.76 exaFLOPS of modelling performance accelerated by AI.
The supercomputer will help tackle many issues in current fusion modeling, including plasma turbulence, reactor material development, and tritium fueling technology advancements.
The system will employe AMD Instinct GPU accelerators and AMD EPYC processors that will run on Dell PowerEdge infrastructure, with WEKA acting as the programme’s storage platform.
Recommended reading
- UK AI Collab with EU To Open Access to Supercomputers
- EU Cuts Ribbon On First Exascale Supercomputer
- New UK-US Partnership Promises ‘Golden Age’ of Nuclear Power
“UKAEA is taking lessons from the Apollo programme: we learn fastest when we can test, iterate, and improve safely in the virtual world before we commit to our real-world mission,” Dr Rob Akers, UKAEA’s director for computing programmes, said.
“Sunrise will bring that capability to fusion by combining high-fidelity simulation with physics-informed AI to develop predictive digital twins that reduce the cost, risk and time of learning that would otherwise require expensive and time-consuming physical testing.”
Sunrise will be used to address real-world challenges from a wide range of UK fusion programmes to drive critical advancements for the LIBRTI (Lithium Breeding Tritium Innovation) programme, which is developing tritium fuel-cycle technologies for self-sufficiency in future fusion operations, and for STEP Fusion, the UK’s flagship initiative to demonstrate fusion energy in the 2040s.





