Why fusion is finally a venture-scale opportunity

For most of its history, fusion has been the punchline of a single joke: it is, and always will be, thirty years away. The joke earned its keep. Six decades of national-lab work produced beautiful physics and, until very recently, no path to a wall-plug-positive machine that anyone would actually build.

That has changed in the last five years, and the change is structural rather than rhetorical.

The three things that crossed

Three independent curves crossed at roughly the same time, and the intersection is what makes the field newly tractable.

Ignition was demonstrated. The National Ignition Facility crossed scientific breakeven in December 2022 and has since repeated the result with rising margins. This is a single facility, on a single approach, and it is not how a power plant will be built — but it ended the open question of whether controlled fusion produces net energy at all. That question is closed.

High-temperature superconducting magnets matured. REBCO tape moved from a laboratory curiosity to an industrial product capable of generating the field strengths that magnetic confinement designs need. This collapsed the size — and therefore the cost — of a viable tokamak by roughly a factor of forty. Commonwealth Fusion Systems’ SPARC, Tokamak Energy’s ST40, and several private machines are direct beneficiaries.

Private capital arrived in serious size. Total private fusion funding crossed seven billion dollars cumulatively in 2024 and continues to grow. This is the first time in the field’s history that the capital available is commensurate with the engineering problem.

What the venture problem actually looks like

The interesting questions in fusion are no longer physics questions. They are operational. The companies that ship working machines this decade will solve a recognisable cluster of supply-chain problems:

• Fuel cycle. Tritium handling, breeding blankets, lithium supply. Almost no one has done this at industrial scale and the regulatory regime is unwritten.
• Magnet supply. REBCO production capacity worldwide is measured in tonnes per year and needs to grow by an order of magnitude.
• Plant engineering. Heat exchangers, vacuum systems, neutron shielding, remote maintenance — every one of these is its own serious business.
• Regulatory pathway. The NRC’s decision to regulate fusion separately from fission was the single most important policy outcome of the decade for the industry.

Each of these is a venture-scale company in its own right. The public attention has focused on the headline machines; the durable investments are likely to live in the picks-and-shovels around them.

What we look at

We watch tokamaks, inertial confinement, and the smaller magnetic-confinement variants without strong priors on which approach reaches commercial generation first — the answer matters less than the supply chain serving all of them. We pay particular attention to component businesses with cross-machine applicability, to fuel-cycle infrastructure, and to the firms quietly building the inspection, simulation, and operating-software layer that any working plant will eventually need.

The thirty-year joke is over. The supply chain is now the question.