Fusion finish line is in sight: Inside ITER’s giant tokamak

Big Tech is investing in fusion. Just how close are we to limitless, safe, carbon-free energy?

In a quiet corner of southern France, an international mega-project is forging ahead with its mission to demonstrate industrial-scale nuclear fusion. Computing got to take a look.

CIOs are increasingly tasked with shaping and driving sustainable digital growth strategies, placing the environmental impact of cloud datacentres—especially those supporting Big AI—under intense scrutiny.

Big Tech companies are already the planets largest purchasers of renewable energy. However, solar and wind power have inherent limitations, prompting the industry’s leading players to diversify their clean energy portfolios by incorporating nuclear fission. This includes signing deals to connect datacentres to existing nuclear reactors and investing in new Small Modular Reactors (SMRs).

Beyond fission, nuclear fusion represents the holy grail for safe, carbon-free energy. Recent advances have brought the prospect of virtually limitless fusion energy tantalizingly close. A good indication of just how close can be discerned from the growing interest of hyperscalers.

In 2023, Microsoft signed an agreement with fusion startup Helion to secure energy supply within five years. Similarly, OpenAI CEO Sam Altman has invested in Helion, and Jeff Bezos backs General Fusion. Altman has emphasised that the future of AI depends fundamentally on the availability of clean energy.

Whilst creating fusion energy is one (let’s be clear, very significant) challenge, making it commercially viable is another. This is the mission of ITER (International Thermonuclear Experimental Reactor), a colossal nuclear fusion research and engineering project involving 33 nations. Located in rural southern France, ITER is building the world’s largest fusion reactor, aiming to demonstrate the integrated systems necessary for industrial-scale fusion. Success would mark a breakthrough in making fusion commercially viable.

Ahead of schedule

ITER, launched in 2006, has faced significant delays, with the project now scheduled to begin operations in 2034—nine years later than initially planned—and energy-producing fusion reactions expected by 2039. These delays stem from political complexities, technical challenges, safety modifications demanded by French nuclear regulators, manufacturing issues, the COVID-19 pandemic, and the unprecedented scale and complexity of the project.

The treaty underlying ITER necessitates coordinating unanimous decisions among diverse international partners, including the US, China, Russia, Japan, India, South Korea, and the EU. This is not an environment for agility or where minimum viable product will do.

Despite setbacks, recent developments indicate that ITER is back on track. Component manufacturing and transportation are concluding, and assembly is well underway. In April 2025, the first of nine sector modules of the tokamak fusion reactor was installed ahead of schedule. These modules will form the doughnut-shaped vacuum vessel where “first plasma” will be generated. Another module is set for installation in July, and six of the required nine are now onsite.

Additionally, the final component of the central solenoid—the reactor’s most powerful magnet, capable of lifting an aircraft carrier—has been tested and is ready for assembly.

Accuracy at scale

The scale of ITER and its tokamak is jaw dropping. The site spans 180 hectares, with the tokamak complex rising 80 meters tall and weighing 23,000 tonnes. The weight of some components has necessitated road modifications between port and site.

For example, the sector module installed in April, which is partially shown in the image above, is as tall as a five-story building and weighs as much as four fully loaded jumbo jets. Components are manufactured across member countries and brought to France for assembly. (The treaty specifies that ITER’s member nations share costs proportionally and contribute in-kind through components and systems, with experimental results and intellectual property also shared.)

Precision is paramount. These vast, complex and terrifyingly expensive modules must be assembled with sub-millimetre accuracy. ITER Chief Scientist, Alain Becoulet likens it to building a Swiss watch but on a massive scale. Magnets, central to the tokamak’s function, must align within tight tolerances despite their enormous size—17 meters tall and weighing hundreds of tonnes. Quality inspection and early detection of non-conformities are critical to controlling costs and ensuring success.

Metrology—the science of measurement—is integral to ITER’s progress, so integral in fact that it requires a team to manage and coordinate it. That team is run by Beatrix Alix - Metrology and Reverse Engineering Coordinator. She explains:

“In the metrology function, we oversee the dimensional risk. So, anything that can happen in terms of dimension for the building of the machine, installation, during the operation and during the maintenance.”

Alix continues:

“It would be impossible to construct ITER without extremely precise methodology. Metrology is everywhere throughout assembly.”

Laser trackers, 3D scanners, and spatial analysis software made by digital reality specialist Hexagon are extensively used by Alix’s team to model and guide assembly processes.

Making fusion viable

As assembly advances, ITER is also focusing on digitisation to manage the vast knowledge, data, and information generated. Alain Becoulet emphasises the importance of being fully operational from day one, leveraging digital tools to optimise operations, maintenance, repairs, future upgrades, and eventual decommissioning. Human resource continuity is another concern given the long-time scale of the project, with digital support playing an increasing role in knowledge transfer to new team members.

How close are we to a world where clean fusion energy makes a up a significant part of the energy mix that industry, especially tech, depends on?

Becoulet does not expect fusion energy to contribute to 2030 net-zero targets. However, he is optimistic about the following decade.

“We want to accompany the energy transformation of the world in the coming decades,” he says. “ITER is probably the last public endeavour in research and development with the mission of demonstrating the feasibility of mastering fusion reaction, but it has no industrial aspect.”

That aspect must develop in parallel with the tokamak construction if ITER is to fulfil its mission goals. The good news is that the private sector is very much interested. A corridor in the ITER complex displays hundreds of plaques bearing corporate logos. ITER invites partners in twice a year – the last meeting attracted approximately 350 companies. Becoulet continues:

“ITER is successful and is developing and is triggering a lot of interest from the private sector. It looks feasible, we can start the industrialisation of this process. And the industrialisation, by definition, is built by industry.

“Every single day at ITER we are transferring information to people who are willing to embark on the industrialisation of fusion.”