Fusion just got its first public stock. The physics still runs on its own clock.

On Friday morning a fusion company rang the opening bell at the Nasdaq, and for the first time in the history of the field, you could buy the stock. General Fusion, a Canadian outfit founded in 2002 and based outside Vancouver, began trading this week under the ticker GFUZ, becoming the first publicly listed pure-play fusion energy company anywhere in the world. "We are honored to ring the Nasdaq Opening Bell and mark this milestone," its chief executive, Greg Twinney, said, "and to celebrate becoming the first publicly listed fusion company." It is a genuine milestone, and I want to be precise about what kind. It is a milestone in capital formation. It is not, yet, a milestone in physics, and the distance between those two things is the entire story.

I keep a private archive of fusion timelines — every "fusion by [year]" and the year it actually turned out to be — and I have learned to read a company by the gap between what it has demonstrated and what it has announced. General Fusion has now announced something no fusion company has done before: it has invited the general public to own a piece of the wait. Whether that is a good idea depends entirely on how honestly the wait is described, so let me try to describe it the way the physics does, rather than the way a bell-ringing does.

What the machine actually does

General Fusion's approach is called magnetized target fusion, and it is worth understanding because it is neither of the two you usually read about. It is not a tokamak, holding a plasma in a magnetic cage for long stretches the way the giant ITER project does, and it is not laser-driven inertial confinement, the approach that achieved a brief net-energy pulse at the U.S. National Ignition Facility in 2022. It sits in between. You form a donut of magnetized plasma — hot, ionized gas threaded with its own magnetic field so it holds together for a moment — and then you crush it, fast, with a physical liner. In General Fusion's design the liner is liquid metal, driven inward by an array of pneumatic pistons that slam closed like the world's most expensive combustion engine. Compress the plasma hard enough and quick enough and, in principle, it reaches the temperature and density at which hydrogen nuclei fuse and release energy. The appeal of the scheme is that the liquid-metal wall also absorbs the neutrons and the heat, sidestepping some of the materials problems that make the tokamak's first wall such a punishing engineering assignment.

The company has been at this for more than two decades. It has run, by its own count, over 200,000 plasma experiments; it holds something like 210 patents issued and pending; it has raised more than $400 million in private capital across its life. Its current machine, Lawson Machine 26 — LM26 — is the one meant to prove the concept at what the company calls a commercially relevant scale. And in June it reported a real result: an electron temperature of 0.72 kiloelectronvolts, which is about 8.4 million degrees Celsius. That is a hot plasma by any honest measure, and reaching it in a machine of this type is a legitimate piece of experimental work. I don't want to wave it away. I want to put it on the correct axis.

On what scale, and how far along it

Here is the axis. Fusion requires three things at once, and the trap in every fusion headline is that it reports one of them and lets you assume the other two. You need temperature, so the nuclei are moving fast enough to fuse. You need density, so they are close enough together to collide. And you need confinement time, so the hot, dense state holds long enough for more energy to come out than you spent getting there. The product of those three — roughly, temperature times density times time — is what has to cross a threshold physicists call the Lawson criterion. Hitting any one of them is not the same as hitting the product, and the machine is literally named for the criterion it has not yet met.

Fusion requires temperature, density and confinement time all at once. The trap in every headline is that it reports one of them and lets you assume the other two.

So take the temperature the company just announced and read it against its own stated roadmap. LM26 has reached 0.72 keV. The company's near-term goal is to reach 1 keV, and then to push toward 10 keV, and then — separately — to demonstrate the Lawson criterion, the actual net-energy condition. In round numbers, the plasma is about a fourteenth of the way up the temperature ladder the company itself has drawn, and temperature is only one of the three rungs. Getting to 8.4 million degrees is real, demonstrated, on the record. Getting to roughly 120 million degrees while simultaneously holding the density and the confinement time long enough to cross breakeven is announced. The company targets those milestones by 2028, and a first commercial power plant, on its own telling, no earlier than the mid-2030s. I have written the phrase "mid-2030s" in this beat enough times to know it is a horizon that tends to recede as you approach it, and I would ask the same question I always ask, out loud: mid-2030s measured from when?

None of this means the science is fake or the company is unserious. It means the machine is early, in the specific way fusion machines are always early — closer than they were, much further than the announcement makes it sound. A reactor that reaches a fusion-relevant temperature in a single shot has demonstrated something real; a plant that reaches net energy reliably, cheaply, and continuously for thirty years is a different problem by orders of magnitude, most of them engineering rather than physics. Conflating the two is how "we made a hot plasma" becomes "fusion is here," and it is precisely the conflation a stock ticker invites.

The number the public actually voted on

Because there is a second set of measurements this week, and they are financial, and they are unusually candid about what investors think of the timeline. General Fusion went public not through a traditional offering but by merging with a special-purpose acquisition company, Spring Valley Acquisition Corp. III. That vehicle had raised about $230 million from investors who were given a standard right: when the merger came up for a vote, they could take their money back instead. About 87 percent of them did. Roughly $30 million of the original trust remained. The company filled the hole with a $107.7 million private placement and some existing cash, ending with around $150 million to work with — real money, but well short of the roughly $314 million the deal originally targeted, and a rounding error against what building a fusion power plant will eventually cost.

I dwell on the 87 percent because it is the most honest number in the whole event. Given the first real chance to price fusion's timeline, the people who already held the shares overwhelmingly chose the exit. The headline valuation — a pro-forma equity value of about $1 billion — sits oddly next to an actual market capitalization closer to $280 million and an annual operating cost around $24 million and no revenue at all. Twinney has been refreshingly plain about the pre-revenue reality. "There's examples of pre-revenue companies" going public, he said before the listing; "it's not typical … but it's also not something that's never been done." His stated reason for going first is access — "to open up this opportunity to invest in fusion to a broader set of investors," meaning the public. That is a real argument. It is also worth naming what it transfers. Private fusion investors are patient capital that understands a payoff arriving after they do. Public markets price on the quarter. The first fusion IPO does not shorten the physics; it moves the timescale mismatch onto a shareholder base whose instrument for expressing doubt is a sell order on a Tuesday.

Two boats, one wave

The listing did not happen in a vacuum, and the surrounding weather cuts both ways. By one industry tally, private fusion has now attracted more than $14 billion in total investment, with a record $4.48 billion in the year to July 2026 across some 56 companies. That is a field with genuine momentum, and General Fusion is a credible member of it, with a serious science advisory bench — its board includes Tony Donné, formerly of the European EUROfusion consortium — and a framework agreement signed last month with the Italian energy firm Renexia to explore eventual deployment. A framework agreement is not a power purchase, and "explore eventual deployment" is doing a great deal of work in that sentence, but it is not nothing, either. The interest in firm, carbon-free power is real, and it is what is funding the patience the physics requires.

What the listing changes is who holds the risk of the gap. I would credit General Fusion for one thing without reservation: it has not, this week, dressed 8.4 million degrees up as an imminent power plant. Its own roadmap lays out the ladder honestly enough that a careful reader can see how many rungs remain. The problem is that a bell-ringing is not a careful reader. It is a headline, and the headline says "first fusion stock," and a great many people will buy the headline without ever finding the 0.72. So here is the sentence I would tape to the monitor of anyone tempted. The company has demonstrated a hot plasma and announced a power source, and between those two words sit a factor of roughly fourteen in temperature, two physical quantities it has not yet shown it can hold at the same time, and a decade the field has a long and documented habit of spending twice. The stock will trade every day this year. The neutrons keep their own schedule, and it has never once been the market's.