Most fusion research happens inside cavernous facilities lined with massive magnets or arrays of enormous lasers. But a startup in Cambridge, Massachusetts, has been quietly building a fusion device that fits on a tabletop — and it just raised $6.2 million to scale it up.

Cambridge Heavy Industries, co-founded by CEO Yun Kai, has been operating in stealth out of a small lab near MIT. The company recently shared an exclusive look at its technology, revealing a benchtop fusion reactor that uses lasers to compress fuel targets — a fundamentally different approach from the giant fusion experiments that dominate headlines.

A Fusion Device Small Enough to Fit on a Bench

"We purposely built a small modular reactor. It's tiny, it's benchtop," Kai explained. "Because it's small, it makes everything easier."

That small size isn't just about convenience. It allows Cambridge Heavy to experiment and iterate at a pace that massive facilities simply can't match. Kai said the device can run 50 to 80 experiments — or "shots" — per day. Each shot generates data that feeds into simulations powered by artificial intelligence, which the team uses to continuously optimize the design.

The concept originated from the lab of Keith Nelson, a chemist at MIT. Nelson's researchers were using laser pulses to study materials under extreme conditions. As they pushed pressures higher, they realized they were approaching conditions that could trigger fusion. That insight eventually became the foundation for Cambridge Heavy's approach.

How the Tabletop Reactor Works

Cambridge Heavy's device uses lasers to compress a fuel target, similar in principle to the groundbreaking experiment that proved controlled fusion could produce more power than it consumes. But unlike other inertial confinement designs, Cambridge Heavy's approach minimizes the number of lasers required.

The system uses two main lasers to drive ignition. One of those lasers is split into multiple beam lines that deliver light in concentric rings. Those rings of laser light are timed to hit the flat, round fuel target successively, starting from the outside and working inward.

Here's the key detail: the first ring starts a compression wave within the target that begins traveling inward. The subsequent laser rings are timed to boost that wave at just the right moment, further compressing the fuel. As the shock wave reaches the center of the target, a laser fires on the target, generating heat and producing a magnetic field. With the right combination of pressure and heat, fusion occurs.

Kai said the design can generate high pressures with relatively low laser power — a key advantage that makes the benchtop form factor possible.

Why This Approach Matters

The significance of Cambridge Heavy's approach goes beyond the novelty of a tabletop fusion device. Two things stand out.

First, the small form factor enables an iterative development cycle that massive fusion facilities can't match. With 50 to 80 experimental shots per day, the company can gather data and refine its design far faster than projects that can only fire a handful of times per year. That data feeds AI-powered simulations, creating a design optimization loop that accelerates progress.

Second, the tritium market represents a real bottleneck for the entire fusion industry. With only about 40 kilograms of tritium in existence globally and a projected 28-kilogram shortfall by 2054, the availability of fusion fuel is not a theoretical problem — it's a concrete supply chain constraint. Cambridge Heavy's plan to produce tritium with its smaller devices addresses this gap directly.

A Clever Business Strategy: Make Tritium First

Before Cambridge Heavy builds a full power plant, it plans to use its smaller-scale devices for something more immediately profitable: making tritium.

Tritium is a rare isotope of hydrogen that nearly all fusion power plants under development today plan to use as fuel. It's scarce — only about 40 kilograms exist globally — and it's expensive, selling for around $30,000 per gram. One recent study predicts that without new sources of tritium, the fusion power industry will face a shortfall of about 28 kilograms by 2054. At current prices, that represents roughly $840 million worth of missing fuel.

Fusion startups are working on fuel systems that would let reactors produce their own tritium, making them largely self-sufficient. But Kai said that process could take up to a decade. In the meantime, Cambridge Heavy's tritium-making devices could fill the gap — giving the company years of potential revenue while it pursues its longer-term power plant goal.

What Happens Next

In about a decade, Cambridge Heavy plans to develop a power plant from its benchtop concept. That will require scaling up the design — it will be larger than the current prototype, though Kai said it should be roughly the size of two basketball courts, far smaller than many other fusion power plants under construction.

The $6.6 million seed round, led by Azolla Ventures with participation from Overlay, Neta Foundation, P3A, MassCEC, and the Rothrock family, will fund that journey. The funding and the growing investor interest suggest that despite fusion's reputation as a decades-away technology, some investors are betting that smaller, faster-iterating approaches may reach the market sooner than expected.

Whether Cambridge Heavy's timeline holds remains to be seen. But the company is at least pursuing fusion with a pragmatic, step-by-step strategy: build a small, fast-iterating device; use it to generate revenue by making tritium; and only then scale up to a power plant. In an industry known for ambitious timelines and long horizons, that approach stands out.

VANTIX Takeaway

Cambridge Heavy Industries is taking a pragmatic approach to fusion — starting with a benchtop device that can iterate quickly, using it to produce tritium for the fusion industry in the near term, and scaling up to a power plant only after proving the concept. The tritium supply gap is real, and the company's near-term revenue strategy is grounded in a genuine market constraint. Whether the power plant timeline holds is the bigger question.