The three high-tech lunar launches fuel SpaceX’s impressive initial public offering

SpaceX will go on the market on Friday, and investors can barely contain their excitement. The stock offering is said to be worth $75 billion Over-subscribe deeplywith some institutional investors racing in Blocks worth $10 billion From Elon Musk’s empire.

There are plenty of reasons to be skeptical of the investment — big IPOs tend to go under, the company loses money, and Musk’s erratic online behavior would be terrifying coming from any other tech CEO — but that doesn’t seem to be slowing anyone down. Technology investors have learned to never bet against Elon, whatever the business logic.

But an honest look at SpaceX’s financial plans can still tell us a lot about what they’re betting on: a company centered around orbital data centers that has emerged in the past 18 months under Musk’s name. He sought to see It would consolidate its group before the IPO.

In true Musk style, this is an audacious scheme, requiring at least three near-impossible feats of engineering: a reusable rocket, a brand-new American chip foundry, and a sprint to build satellites faster than ever before.

This type of business plan can be difficult to achieve. This week, two analyzes attempted to provide a more realistic assessment of SpaceX’s plan — one from financial research firm Morningstar, and the other from Aswath Damodaran, a finance professor at New York University who is particularly interested in valuing companies. Both trades found SpaceX to be worth far less than the roughly $1.8 trillion valuation given by the company’s bankers. Morningstar Sets a value By about $825 billion, while Damodaran suggests The company is worth $1.2 trillion.

The big difference is, in many respects, the result of shifting the global space monopoly to the riskier AI business. The Morningstar analyst chalks the difference between their fair value assessment of $63 per share, and SpaceX’s offer price of $135, as a $72 call option on the company’s ability to deliver orbital data centers at the rate and capacity Musk believes is possible.

In both analyses, the high margins of the company’s space launch and satellite Internet businesses are the most attractive things for the company, while its artificial intelligence businesses are the most obscure.

To cloud or not to cloud?

Part of the question is, what does SpaceX’s AI do? In the company S-1 Analyzing the market, it frames its biggest opportunity in enterprise AI — that its models will power programming tools built by the team it acquired from Cursor, or the company’s Macrohard project, which aims to provide digital agents with the capabilities needed to perform white-collar work. SpaceX estimated the total market for this work at $22.7 trillion, compared to $2.4 trillion for AI infrastructure and just under $2 trillion for the company’s space efforts.

But this contradicts recent deals the company has made to sell large amounts of its computing Anthropic and Googlevirtual competitors in the typical business. This is not out of place for Musk; SpaceX frequently launches satellites operated by competitors to its Starlink network. He usually does it from a place of strength, not while playing catch-up.

Acting like the new cloud may be good business in the near term, but it raises the question of where value will accrue in the AI ​​technology stack: Is it better to be a compute provider or a model builder, if you can’t do both?

The scaling logic that dominates AI work requires that serious frontier labs constantly train new and more powerful models (or as Musk put it: I confess In his recent lawsuit against Sam Altman, by extracting capabilities from other companies’ models). Any competitor that does not push forward is likely to fall behind, although the increasing capabilities of cheaper open source models may undermine this dynamic.

Space data centers are one way to square the circle, providing so much compute that SpaceX can effectively do both.

Musk’s space data center architecture

In a Video interview Released by SpaceX this week, Musk laid out the logic why SpaceX is better positioned to deliver its services in data centers. The crux of the argument was that SpaceX is the only company that can put a lot of mass into orbit cheaply, build a lot of solar panels, and build a lot of chips. In general, industry experts see large-scale space-based data centers as being about a decade away, but Musk has said (with many caveats) that it’s much closer.

“This is not a promise of what we will do,” Musk said in the video. “That’s what we’re going to try to do, and I think we can probably do, is get to an annual rate of roughly a gigawatt a year by the end of next year, in terms of space AI computing.”

Based on an expected maximum power delivery of 150 kilowatts per satellite, this means a production rate of 6,666 satellites per year, or about 556 satellites per month. This is nearly double the current advertised production rate of Starlink satellites, which is just 70 per week. Although Musk says the AI ​​satellites are simpler in architecture, that’s a lot to ask for a production facility that hasn’t been built yet. The company is also still building a solar panel production facility.

And that’s before we get to Terafab, the company’s much-discussed chip foundry, which Musk sees feeding later stages of this product as the company tries to ramp up its annual compute output to terawatts. Chip factories are some of the most difficult modern industrial projects, typically costing billions of dollars and taking up to a decade to build.

Then there’s the more important question: What about the spacecraft, the key to SpaceX’s ability to economically put all those chips into orbit?

A recent test flight went well enough, but it didn’t indicate that the possibility of rapid reusability was anywhere near. SpaceX may end up being reused Just the supporter Initially, this would raise startup costs for a space data center. Currently, the company is still the subject of an unfortunate FAA investigation to understand why the booster stage failed to perform a controlled reentry as planned. SpaceX did not respond to questions about when the vehicle would fly again, saying it expects to start launching Starlink satellites with it by the end of this year.

But take that with a grain of salt: Keep in mind that NASA, which has a roughly $4 billion contract with SpaceX to use Starship as a lunar lander, is still not ready to commit to a test mission with the vehicle scheduled for late 2027.

Buyer beware

When public investors get their hands on SpaceX shares, they will find themselves owning a near-monopoly on access to space in the United States and Europe, a global communications network, and a bet on the most ambitious infrastructure project of the age of artificial intelligence.

These projects depend on SpaceX creating something never before seen: a fully reusable rocket. The company will also need to build a high-speed production facility for the AI ​​satellites, but do so within 18 months, not the decade it took to develop Starlink manufacturing. Finally, it would need to build a chip foundry in the United States, something even silicon companies are reluctant to do. Musk is right that SpaceX is the only company capable of building any of this anytime soon, but that says as much about the scale of the challenge as the likelihood of the company achieving it.

Musk used to say he wouldn’t take SpaceX public until it got to Mars, because skittish investors might lose confidence along the way. Those plans may have been postponed, but what comes before the company’s IPO could be just as difficult.