Commonwealth Fusion Systems Gets $85 Million to Build Fusion-Ready Semiconductors

Commonwealth Fusion Systems said it will receive $85 million from the U.S. government to develop semiconductors for the fusion power industry. The funding is a CHIPS Act award from the Department of Commerce, and under the award the U.S. government will take a stake in the company.

The award is a proposal that has not been finalized. Commonwealth Fusion Systems signed a letter of intent with the Department of Commerce for the proposed $85 million, according to the company's announcement on X and coverage on Yahoo Finance. The stated purpose is to accelerate next-generation semiconductor research and development and to boost U.S. R&D and manufacturing capacity for chips.

The company has organized the money into five initiatives. Four of them have been described in detail:

• Advancing silicon carbide power electronics, a wide-bandgap semiconductor material that tolerates high voltage and heat with lower losses.
• Developing next-generation radio-frequency (RF) power generators that heat fusion fuel.
• Building advanced electronics for faster reactor control that are hardened against radiation.
• Adapting AI hardware, including GPUs and TPUs, for use in power plants.

TechCrunch reported the breakdown on October 23, 2026.

Commonwealth Fusion Systems builds tokamak fusion systems, which confine superheated plasma with magnetic fields, around high-temperature superconducting magnets developed in collaboration with MIT. The company says those magnets allow smaller and lower-cost machines. Brandon Sorbom, co-founder and chief science officer, is the company's most senior technical figure named in the announcement.

For engineers, the list reads as a set of concrete problems. RF heating chains must deliver power reliably into a plasma. Control electronics must respond quickly enough to manage a reactor, and they must survive a neutron and gamma environment that conventional parts were never designed for. Silicon carbide devices are attractive because they handle high voltage and temperature with lower losses, which matters when a plant draws power at scale. Adapting GPUs and TPUs points to running inference-heavy control and diagnostic workloads close to the machine, rather than in a distant data center.

In my view, the more significant question is whether a federal chips program will fund the unglamorous electronics that determine whether fusion plants can be operated, not only the physics that gets them started. Much of the public discussion of fusion concentrates on plasma confinement and magnet performance. The electronics around those systems have received far less attention, and an award of this kind puts money behind that layer.

The broader context here is the government's equity position. A CHIPS Act award that carries a stake in the recipient is a structure that the program's critics and supporters will both watch. The letter of intent means the final agreement may differ. Readers who follow semiconductor policy will want to see the definitive documents before drawing conclusions about the precedent.

What the announcement does establish is a specific technical agenda. If radiation-hardened, fast-control electronics and efficient RF generation can be made reliable, the constraints on operating a fusion plant change, even though the physics question is separate. That is the outcome the company is pursuing, and the next public milestones will show whether the work moves from proposal to hardware.