Cambridge Heavy Industries is building a tabletop fusion reactor

Fusion power is usually associated with room-sized facilities built around massive lasers or magnets. Cambridge Heavy Industries has taken the opposite approach: a small, modular reactor compact enough to sit on a lab bench. Co-founder and CEO Yun Kai told TechCrunch the small size lets the company iterate fast, running 50 to 80 experimental “shots” per day, with data from those runs feeding AI-driven simulations that optimize the design.

The reactor uses lasers to compress a fuel target, the same basic approach — inertial confinement — that produced the first controlled fusion reaction to generate more energy than it consumed. Cambridge Heavy’s version uses fewer lasers than other inertial-confinement designs. One laser splits into multiple beam lines that fire concentric rings of light at the flat, round fuel target in sequence, from the outside in, each ring timed to boost a compression wave already traveling toward the center. When the shock wave reaches the target’s core, a separate laser fires, generating heat and a magnetic field that triggers fusion under the right pressure and temperature. The concept originated in the MIT lab of chemist Keith Nelson, where researchers studying materials under extreme pressure with laser pulses realized they were nearing fusion-ready conditions.

The seed round was led by Azolla Ventures, with Overlay, Neta Foundation, P3A, MassCEC, and the Rothrock family also participating. Cambridge Heavy plans to spend roughly a decade scaling the concept into a power plant — one Kai says should be about the size of two basketball courts, smaller than other fusion plants — to test whether the design can generate enough energy to support it. Before that, the company plans to use its smaller devices to produce tritium, a rare hydrogen isotope nearly all fusion plants in development will need and which currently sells for around $30,000 per gram. Only about 40 kilograms of tritium exist today, and one study predicts the fusion industry would be short by about 28 kilograms by 2054 — worth roughly $840 million at current prices — without new sources. Fusion startups are working on fuel systems to make reactors largely self-sufficient on tritium, a process Kai said could take up to 10 years, leaving Cambridge Heavy a window to sell tritium as revenue while it pursues its power-plant ambitions.

Why it matters

A fusion startup selling a scarce isotope as a near-term business, rather than waiting on a power plant that is a decade out, gives it revenue before the harder bet pays off — a model other fusion ventures may need to copy as tritium supply tightens.