One news headline this week had a whiff of déjà vu about it. Nuclear startup Deep Fission announced that it was going public, hoping to garner investor support to build subterranean reactors to power AI data centers.

Wait, didn’t I already write that story? I could have sworn that I did. 

Oh right, I did. Last September, Deep Fission said that it had gone public via a reverse merger with Surfside Acquisition, a Delaware shell company, a transaction in which a private company acquires an existing publicly listed entity to gain a stock market listing — raising $30 million in a concurrent private placement at $3 a share. Now it’s seeking $157 million in a Nasdaq IPO at $24 to $26 a share. You can see my confusion.

Turns out the previous public listing was public in name only. The reverse merger with Surfside was completed, making Deep Fission a reporting company with SEC obligations, but its stock never actually traded. The company had said it intended to list on the OTCQB, a marketplace for developing companies that don’t meet the listing requirements of major exchanges like the NYSE or Nasdaq. But searches for Deep Fission on OTCQB don’t return any results, and the company, in its S-1, denied that its stock had ever been publicly traded.

In response to questions from TechCrunch, Deep Fission declined to comment, citing the quiet period before its IPO.

Deep Fission’s new public offering on Nasdaq is following the more traditional IPO route, with an offering that would value the company at up to $1.66 billion. It’s a sizable figure for a company that one year ago was struggling to raise a $15 million funding round.

Stranger still, the picture painted in the S-1 filed on May 20 is arguably bleaker than the one outlined in the December filing with the SEC. Its timeline for turning on its first reactor has slipped. Further, back in December, it had hoped to achieve criticality — the point at which a nuclear chain reaction becomes self-sustaining — by July 2026. Now, it won’t provide an estimate. 

Deep Fission does point out that it is drilling a test well. It has also lost a lot of money. 

One thing that hasn’t changed: The new S-1 statement contains the same “going concern” warning present in December. If Deep Fission doesn’t complete the IPO, it could run out of money in the next 12 months.

In fact, the startup’s financial position has worsened in recent months. As of March, its deficit had grown to $88.1 million from $56.2 million. In the last month and a half, the company’s cash and cash equivalents declined by $6.4 million, or about 7%.

On the technical front, Deep Fission says it is now prioritizing drilling, perhaps a tacit admission that making holes in the ground isn’t as easy as it sounds. 

The company says it started drilling the first of three test wells in March. The well will be used to collect data “up to 6,000 feet deep.” At eight inches in diameter, it’s quite a bit smaller than will be needed at commercial scale. 

The challenges in moving from a test well to commercial scale are likely to be significant. Deep Fission says it will need boreholes 30 to 50 inches in diameter and a mile deep, though it hasn’t settled on a specific dimension yet. Even at the low end, its boreholes will be larger than what’s typically used in the oil and gas industry. And until Deep Fission knows how large of a hole it can drill, it’ll have a hard time finalizing its reactor design.

So what has changed since December that would spur a bigger offering at a nine-figure valuation? The company did receive an $80 million equity investment, including $20 million from data center developer Blue Owl, which also signed a non-binding MOU for future power plants. Still, that wasn’t enough to stave off the going concern warning. It’s possible that Deep Fission is sitting on some positive information that it omitted from the S-1, though that’s hard to believe given what’s riding on the IPO.

It’s more likely that the company and its backers are seeking to capitalize on investor excitement over fission power. Just last month, nuclear fission startup X-energy went public in an upsized IPO. But unlike Deep Fission, X-energy is generating revenue and is significantly farther along in the Nuclear Regulatory Commission’s licensing process — a contrast that serves as a useful reminder that in a sector where enthusiasm can run well ahead of technical and regulatory reality, valuation and progress aren’t the same thing.

It isn’t exactly clear what factors are driving Deep Fission toward its IPO, but technological or commercial progress doesn’t seem to be among them.