UK Invests Sizeable in AI Supercomputer ‘Sunrise’ to Accelerate Fusion Energy
The UK government is making a significant bet on the future of clean energy, investing £45 million in ‘Sunrise,’ a 1.4MW AI supercomputer dedicated to accelerating fusion energy research. Slated to begin operation in June, Sunrise will be located at the UK Atomic Energy Authority’s (UKAEA) Culham Campus in Oxfordshire and is positioned as a cornerstone of the UK’s first AI Growth Zone.
The Promise of AI-Powered Fusion
Fusion research has historically relied on extensive physical experiments and large-scale simulations to understand the complex behavior of plasma and reactor materials. Sunrise aims to revolutionize this process by combining high-performance computing with physics-informed AI models. This will enable researchers to create detailed simulations and “digital twins” of fusion systems, reducing the need for costly and time-consuming physical testing.
Dr. Rob Akers, director of computing programs at the UKAEA, described the approach as an “Apollo program” style for fusion development, allowing for virtual testing and refinement before real-world construction. This shift towards predictive modeling is expected to significantly reduce both the cost and the time required to achieve viable fusion energy.
Technical Specifications and Capabilities
Sunrise is expected to deliver up to 6.76 exaFLOPS of AI-accelerated modeling performance. Whereas this figure specifically relates to AI workloads, it represents a substantial increase in the UK’s fusion research capabilities. The system will incorporate AMD EPYC processors, AMD Instinct GPU accelerators and Dell PowerEdge infrastructure, with WEKA providing the storage platform. Intel and the University of Cambridge are also contributing to the project.
Supporting Key UK Fusion Initiatives
The supercomputer will directly support several crucial UK fusion programs, including the LIBRTI program, focused on tritium fuel-cycle technologies, and the government’s STEP project. STEP aims to build a prototype spherical tokamak power plant in Nottinghamshire, with the goal of demonstrating fusion power generation by the 2040s.
A Broader Push for AI and Supercomputing
The investment in Sunrise is part of a larger UK government initiative to expand domestic AI and supercomputing capacity. A separate £36 million investment has been confirmed for the Cambridge supercomputing center, and Culham is expected to turn into a central hub for AI-driven scientific computing related to energy research.
Challenges and the Path Forward
While the potential benefits of AI in fusion research are significant, challenges remain. Achieving commercial fusion power is a notoriously difficult undertaking. But, the UK’s investment in Sunrise reflects a belief that increased computing power and advanced modeling techniques can accelerate progress towards this goal.
Frequently Asked Questions
What is fusion energy? Fusion is a process that powers the sun and stars, where light atomic nuclei combine to form heavier nuclei, releasing vast amounts of energy. It offers the potential for a clean, sustainable, and virtually limitless energy source.
What is an exaFLOP? An exaFLOP is a measure of a computer’s performance, representing one quintillion (1018) floating-point operations per second. It’s a key metric for evaluating the capabilities of supercomputers.
What is a digital twin? A digital twin is a virtual representation of a physical object or system. In the context of fusion research, digital twins will allow scientists to simulate and analyze the behavior of fusion reactors without the need for physical prototypes.
What is the STEP project? STEP (Spherical Tokamak for Energy Production) is a UK government-funded project to design and build a prototype fusion power plant, aiming to demonstrate the feasibility of commercial fusion energy.
Where will Sunrise be located? Sunrise will be located at the UK Atomic Energy Authority’s (UKAEA) Culham Campus in Oxfordshire.
When is Sunrise expected to be operational? Sunrise is targeted for operation in June of this year.
