Commonwealth Fusion Systems installs reactor magnet, lands deal with Nvidia

Commonwealth Fusion Systems (CFS) said on Tuesday at CES 2026 that it had installed the first magnet in its Sparc fusion reactor, the demonstration device that it hopes to turn on next year.

The magnet is the first of 18 that, when the reactor is complete, will create a doughnut-like shape that will produce a powerful magnetic field to confine and compress superheated plasma. If all goes well, that plasma will release more energy than it takes to heat and compress it.

After decades of promise and delay, fusion power appears to be just around the corner — CFS and its competitors are locked in a race to deliver the first electrons to the grid sometime in the early 2030s. If it pans out, fusion power could unlock nearly limitless clean energy in a package that resembles a traditional power plant.

Key components of Sparc’s magnets have been completed, and the company expects to install all 18 by the end of the summer, said Bob Mumgaard, CFS’ co-founder and CEO. “It’ll go bang, bang, bang throughout the first half of this year as we put together this revolutionary technology.”

When installed, the D-shaped magnets will sit upright on a 24-foot wide, 75-ton stainless steel circle known as a cryostat, which was set in place last March. The magnets themselves weigh about 24 tons each and can generate a 20 tesla magnetic field, about 13 times stronger than a typical MRI machine. “It’s the type of magnet that you could use to, like, lift an aircraft carrier,” Mumgaard said.

To hit that strength, the magnets will be cooled to -253˚ C (-423˚ F) so they can safely conduct over 30,000 amps of current. Inside the doughnut, plasma will be burning at more than 100 million degrees C.

To work out as many kinks as possible before Sparc is turned on, CFS said on Tuesday that it is working with Nvidia and Siemens to develop a digital twin of the reactor. Siemens is supplying the design and manufacturing software, which will help the company collect data to feed it into Nvidia’s Omniverse libraries.

That won’t be CFS’ first simulation — the company has already been running numerous simulations to predict the performance of various parts of the reactor — but the existing efforts provide results in isolation, Mumgaard said. With the digital twin, he said, “these are no longer isolated simulations that are just used for design. They’ll be alongside the physical thing the whole way through, and we’ll be constantly comparing them to each other.”

The hope is that CFS can run experiments or tweak parameters in the digital twin before applying them to Sparc itself. “It will run alongside so we can learn from the machine even faster,” he said.

Building Sparc has been a costly endeavor. CFS has raised nearly $3 billion to date, including an $863 million Series B2 round in August that included investments from Nvidia, Google, and nearly three dozen other investors. The company’s first commercial-scale power plant, Arc, will be the first of its kind. As a result, it will likely cost another several billion dollars, CFS estimates. 

Mumgaard hopes that digital twins and AI technology will help the company deliver fusion power to the grid sooner than later. “As the machine learning tools get better, as the representations get more precise, we can see it go even faster, which is good because we have an urgency for fusion to get to the grid,” he said.

Follow along with all of TechCrunch’s coverage of the annual CES conference here.