Votre adresse email ne sera utilisée que dans le cadre de campagnes d'information ITER Organization auxquelles vous êtes abonné. ITER Organization ne communiquera jamais votre adresse email et autres informations personnelles à quiconque ou dans le cadre d'informations commerciales.
Si vous changez d'avis, il vous est possible de vous désinscrire en cliquant sur le lien 'unsubscribe' visible dans vos emails provenant d'ITER Organization.
AI, high-performance computing and fusion: a report
AI, high-performance computing and fusion: a report
A new report from the non-profit environmental organization Clean Air Task Force (CATF) explores how artificial intelligence (AI) and high-performance computing (HPC) are accelerating fusion energy development in key areas, including materials selection, high-temperature superconductors, inertial fusion energy, tritium breeding, and advanced diagnostics.
A Survey of Artificial Intelligence and High Performance Computing Applications to Fusion Commercialization was released as the COP29 climate conference was underway in Azerbaijan and introduced at the ITER booth on 15 November by Sehila Gonzalez de Vicente, CATF Global Director for Fusion Energy.
Key findings from the 119-page report include:
The arrival of new cloud-based computing tools and AI accelerates prospects for commercialized fusion energy
The new ecosystem for high-performance computing and AI will be critical in both how we deliver and use fusion energy
The fusion-tech nexus could deliver revolutionary results in materials science, control systems, and superconductors
From the report's Executive Summary: "Until recently, fusion's rate of progress was set by empirical science: building machines and testing ideas, generating datasets, and then choosing the next design based on empirical scaling. This manual design process could be a years-to-decades-long process. [...] Today, AI and HPC can speed up progress by doing much of the empirical scaling 'in silico.’ HPC-enabled “parameter sweeping,” or determining which design configurations are potentially viable, helps narrow down the nearly infinite range of possible machine parameters. [...] The speed at which these tools work can drastically reduce the time it takes for developers to move from initial exploration to a basic conceptual design suitable for further development. Furthermore, advancements in cloud-based HPC have democratized access to computational tools that were previously limited to dedicated supercomputers. This has broadened access to high-performance computing across the fusion ecosystem, enabling a diverse range of concepts, including less common fusion concepts, to take advantage of modern simulation tools to accelerate the investigation and optimization of their designs. [...] AI and HPC’s ability to change the design process itself could assist with bringing facilities online sooner."
A new video explores how the WEST tokamak, run and operated by the Institute for Magnetic Fusion Research (IRFM) at the French Alternatives Energies and Atomic Energy Commission's Cadarache site near ITER, is helping to prepare for ITER operation.
WEST, for W (tungsten) Environment in Steady-state Tokamak, is the former Tore Supra, transformed with a full-tungsten, actively cooled divertor that is based on the same design and technology as the ITER tungsten divertor. Experiments at WEST allow scientists and engineers to address risks both in terms of industrial-scale manufacturing and in the operation of the components. A joint team made up of representatives of the CEA, the ITER Organization and the European Domestic Agency Fusion for Energy meets regularly to coordinate tests and share the lessons learned from the WEST experiments.
WEST is currently running a campaign of tests. Join the team in the control room here.