The United Kingdom Atomic Energy Authority (UKAEA) and the US Princeton Plasma Physics Laboratory (PPPL) have established a transatlantic connection between their respective supercomputing facilities. This initiative creates a “fusion data federation” designed to accelerate the application of artificial intelligence to key challenges in fusion energy research. The link connects UKAEA’s advanced computing resources at the Culham Science Centre with the high-performance computing systems at PPPL, a U.S. Department of Energy national laboratory. The collaboration aims to provide a framework for researchers in both countries to jointly develop and deploy AI models on shared datasets, streamlining efforts toward commercial fusion. Source: PPPL
A primary objective of the partnership is the development of sophisticated “digital twins” for future fusion power plants. These high-fidelity virtual models are critical for designing and operating complex systems like the UK’s Spherical Tokamak for Energy Production (STEP) prototype plant. By integrating experimental data with predictive AI algorithms, these digital twins will allow for the simulation of plasma behavior, materials performance, and plant-wide operational scenarios. This computational approach is intended to de-risk the design process and optimize the operational efficiency of commercial-scale fusion machines before construction begins. Source: PPPL
A primary objective of the partnership is the development of sophisticated “digital twins” for future fusion power plants.
The agreement builds upon the 2023 US-UK Strategic Partnership, which identified fusion energy as a priority area for bilateral cooperation. According to PPPL Director Steve Cowley, the grand challenges of fusion require the full capacity of exascale computing and AI. The linked infrastructure will support research into plasma control, disruption prediction, and the analysis of vast diagnostic datasets from experiments worldwide. This effort reflects a growing trend across the fusion industry to integrate advanced computation and machine learning directly into the research and development cycle, moving beyond traditional theoretical and experimental methods. Source: PPPL
For both institutions, the collaboration enhances their existing computational capabilities. UKAEA’s supercomputing resources are a core component of the UK’s national advanced computing facility, while PPPL operates some of the most powerful systems dedicated to fusion science within the US national lab complex. Tim Bestwick, UKAEA’s Chief Technology Officer, stated that the partnership is essential for designing and operating future machines. The data federation is expected to establish common standards and tools, enabling more seamless and efficient collaboration on some of the most computationally intensive problems in plasma physics. Source: PPPL