There have already been some high-profile milestones over the past few years. Toward the end of 2022, the National Ignition Facility at Lawrence Livermore National Lab beat breakeven, creating a fusion reaction that produced more energy than it took to heat up the fusion plasma. Or when the startup Commonwealth Fusion Systems, a.k.a. CFS, announced that it had developed a new type of extremely powerful magnet to better contain and control superheated plasma. Now, startups and investors think the next decade will be critical for commercialization.
“When we started BEV, we kind of assumed that fusion was going to be too far off,” said Larochelle. But after talking with CFS and learning more about the company’s magnet tech, minds changed. Breakthrough invested in the company — and eventually three other fusion startups, too. “These better magnets matter a lot,” Larochelle told me. “It matters as much as the transistor did to a computer. It’s that level of component level breakthrough that totally changes the game.”
For the ordinary optimist, fusion energy might invoke a cheerful Jetsons-style future of flying cars and interplanetary colonization. For the cynic, it’s a world-changing moment that’s perpetually 30 years away. But investors, nuclear engineers, and physicists see it as a technology edging ever closer to commercialization and a bipartisan pathway towards both energy security and decarbonization.
To some extent at least, the data backs them up. According to the Fusion Industry Association, over 60% of all private fusion companies were founded in 2019 or later. And in the past three years alone, fusion companies have brought in over $5.1 billion, over 70% of the sector’s total funding since 1992.
“We would hope to see a breakeven moment by private companies in the next two to three years, by 2028-ish,” followed by a commercial reactor in the mid-2030s, Julien Barber, an investor at Emerson Collective, told me. Thus far, Emerson, which is headed by Laurene Powell Jobs, has invested in two fusion companies, CFS and Xcimer Energy.
The major players in the startup ecosystem say they’re on track to get there. “The progress has actually been faster than Moore’s law,” Ally Yost, senior vice president of corporate development at CFS, told me, “but people weren't looking at that.”
Moore’s law is a prediction — largely validated for decades — that the number of transistors on a microchip, and thus a computer’s processing speed, would generally double every two years. The performance of fusion reactors, especially the donut-shaped tokamak reactors that CFS uses, has historically improved at an even faster rate. But due to some midcentury researchers and technology enthusiasts overpromising on the near-term feasibility of fusion, cynicism remains. It also doesn’t help that the large, intergovernmental fusion megaproject known as ITER has consistently faced delays and huge cost overruns due to the technical complexity of the project, as well as the difficulty of wrangling 35 countries to work together.
Thus far, though, the private sector is faring better. CFS has raised over $2 billion, more than any other private company in the space. It uses an approach known as magnetic confinement fusion, which involves using strong magnets to confine fusion fuel in the form of a plasma. If you can keep the plasma dense enough and hot enough for long enough, atoms start fusing together, releasing a vast amount of energy in the process. ITER, as well as startups including Type One Energy, Thea Energy, and Renaissance Fusion are pursuing the same fundamental route, though with their own technical twists.
Lawrence Livermore, on the other hand, achieved its breakthrough fusion reaction (which it’s since repeated several times) using an approach known as inertial confinement, in which powerful lasers fire at a pellet of fusion fuel, causing rapid compression and heating that leads to nuclear fusion. But the national lab is not aiming to create a commercial reactor. So when the founders of the startup Xcimer Energy saw that the National Ignition Facility was closing in on its goal, they jumped to get inertial confinement tech ready for market.
“In August of 2021, NIF achieved a fusion gain of about 0.6,” Xcimer’s President and CTO, Alexander Valys, told me, referring to the ratio of the energy generated by the fusion reaction to the energy required to heat the fusion plasma. An energy gain of one constitutes breakeven, so the moment didn’t get any mainstream press to speak of. “But inside the field, everyone knew that the previous NIF shot record was effectively a gain of like 0.01,” Valys said. The massive jump indicated to him that, “If we’re going to do this, we have to do it now.” Since then Xcimer has gotten backing from the biggest names in the space, including BEV, Lowercarbon Capital, and Emerson Collective, as it looks to build lasers at lower cost and higher power.
One thing that ties fusion’s various technical approaches together is the fact that they’ve all benefited tremendously from advances in supercomputing, which allows researchers to better model plasma physics and rapidly simulate fusion experiments. “It’s really taken the advent of modern computational methods and supercomputers to be able to model that process with sufficient accuracy, that you can actually develop a machine that recreates those conditions,” Christofer Mowry, CEO of the magnetic confinement startup Type One Energy, told me.
At this point, many leading companies say that the problem is no longer about basic science, but cost. Clea Kolster, head of science at Lowercarbon Capital, told me that once CFS turns on its demonstration reactor, the company knows its fusion gain will be “at least greater than two.” (Lowercarbon is a CFS investor.) That said, there’s still loads of uncertainty around the reactor’s performance, as outside studies project that its energy gain will be more like 11 — although even that might not be enough for it to make economic sense.
So while the economics of fusion are a large part of what venture capitalists are betting on these days, private investment in the industry has actually fallen over the past two years, after peaking in 2022 at $2.8 billion. “A step change in growth will be required once private companies deliver results on their prototype machines,” Andrew Holland, CEO of the Fusion Industry Association, said in a statement, adding that last year’s $900 million in funding “will not be enough to deliver fusion’s ambitious goals.”
To date, government funding has comprised a mere 6% of the industry’s total, but contra the private funding trend, that figure has been ticking up as of late. Last year, the Department of Energy announced $46 million in funding for eight private fusion companies to help the administration reach its goal of demonstrating fusion at pilot scale within a decade.
All the companies I spoke with were awardees, and all agreed that much more would be needed, pointing to the public-private partnership between NASA and SpaceX as a model for how the government could more deeply support commercialization of fusion. That partnership was the product of NASA’s Commercial Orbital Transportation Services program, designed to catalyze the development of private spacecraft and funded to the tune of $800 million.
China, meanwhile, is outspending the U.S. on fusion, just as it’s done with solar, and launched a national fusion consortium at the beginning of this year.
“We are about to harness the sun a second time, and we can’t make that mistake again. We have to get serious about building this industry here in the United States,” Clay Dumas, a partner at Lowercarbon Capital, told me. The firm has a dedicated $250 million fusion fund, and has invested in a total of eight companies in the space, spanning a wide array of technical approaches. “That is going to take the combined efforts of investors and entrepreneurs and policymakers and energy companies and governments to make sure that we can drive this forward on the timeframe that it needs to happen.”