Secretive Startup Builds Benchtop Fusion Device Firing 80 Shots a Day

What Cambridge Heavy Industries is actually building

Not a power plant. A compact testbed designed to make fusion fuel fast enough to learn from it.

Cambridge Heavy Industries has been operating in stealth mode from a lab attached to MIT, according to TechCrunch, and is now sharing details of its approach. CEO and co-founder Yun Kai calls the device "tiny" and "benchtop" sized, which is the whole point: small experiments run quickly, and quick experiments iterate.

The long-term target is a reactor roughly the size of two basketball courts rather than a full power plant complex. That is still enormous by any normal standard, but a fraction of conventional laser fusion facilities.

The design uses two lasers rather than one. The first delivers light as concentric rings onto a fuel block, creating shock waves that travel inward and compress the material in an accelerating squeeze. The second laser hits that wave at the right moment to push pressure high enough for ignition. Because everything moves inward, the enclosed system works as its own compact fuel factory. It is also a 2D layout rather than 3D, which means fewer angles to cover. The National Ignition Facility in the US, by comparison, uses an array of 192 lasers arranged in three dimensions to compress its fuel targets.

50 to 80 shots a day versus a handful at NIF

This is the number that matters. The benchtop fusion device can get through 50–80 laser shots per day, according to Kai, while NIF manages just a handful. Each shot needs relatively little laser power, and dozens of experiments every 24 hours means the team can analyze results and adjust far faster than a facility that fires occasionally.

Fusion research has always been slowed by the cost and rarity of experiments. Making iteration cheap is arguably a bigger deal than any single result the device produces.

The tritium problem nobody wants to talk about

Fusion's best fuel mix needs tritium, and there is very little of it around.

Fusion smashes light atoms together, and the favoured pairing is deuterium with tritium. Deuterium is plentiful. Tritium is not. Most of today's supply comes from older fission plants that are being phased out.

The plan is for future fusion reactors to breed their own tritium while producing electricity, but those reactors are not expected online until the 2030s at the earliest. That leaves a window, and Cambridge Heavy Industries wants to fill it.

TechCrunch points to a recent study estimating the industry-wide tritium shortfall could reach 28 kilograms (62 pounds) by 2054. At today's prices that is roughly $840 million worth of the isotope — a market worth chasing before a single fusion watt reaches the grid.

To be clear about what this device is not: it is not doing the job of NIF or any reactor in development. It makes fuel. Electricity generation comes later, and practical fusion power is still years away, as it has been for decades.

What it means for you

Nothing changes on your energy bill yet. Cambridge Heavy Industries has not announced commercial tritium sales, reactor timelines or pricing, and the company is still in stealth mode, so any product is years out.

For US, UK and EU readers, the more relevant signal is where fusion money is flowing. Governments and private investors are funding reactor designs that will all need tritium, and a supply crunch would hit those projects first. Watch for offtake deals or partnerships with fission plant operators as the earliest sign this becomes real.

Fusion has a habit of promising more than it delivers, and this device does not change that. What it does do is attack a boring, concrete bottleneck — fuel supply — with a tool built for fast learning. If tritium really is the choke point, the startup that solves it early gets a seat at the table long before the first fusion plant switches on.