SpaceX’s plan: Put all the data centers in space, then profit

The world’s most valuable companies are almost impossible to avoid. One makes your iPhone. One delivers packages to your door. One built the Windows operating system running in workplaces worldwide, while another makes chips powering countless computers and devices. There’s the firm behind a scrappy search engine you may have heard of: Google.

And then, there’s SpaceX — Elon Musk’s money-losing rocket startup that has soared into the orbit of all of those tech juggernauts with a fleet of fantastical plans instead of ubiquitous products. With a more than $2.4 trillion valuation as of Thursday, SpaceX is the world’s sixth-most-valuable company.

Its market value is based on promises of putting data centers in space and establishing a colony of a million people on Mars, a premise so far-fetched that three experts reached by The Washington Post to break down the factors driving SpaceX’s valuation declined to ascribe any impact to it, despite the fact that Musk’s compensation is tied to SpaceX achieving that goal.

“[If] anyone else said it, they would probably have him institutionalized,” said Greg Martin, managing director of private markets at the firm Rainmaker Securities, though he noted Musk has delivered on ambitious promises in the past.

SpaceX has joined the ranks of the major forces in tech after making soaring promises about its potential to dominate the emerging space economy. The company, which has lost $13 billion since the start of 2023, aims to become a hub for computing power overhead — fueled by the sun and cooled in the frigid vastness of space, an idea that has led to skepticism.

Putting data centers in space “was always going to be” the solution, Musk said Thursday, “as you can scale a trillion times more than you could on Earth.”

Despite Musk’s record of overpromising, launching a fleet of data centers into orbit is being taken seriously in some corners.

Though SpaceX lacks even a proof of concept for a space-based data center, companies including Anthropic and Google have made deals with it to rent out its data center capacity on Earth, the beginnings of a line of business the company hopes to vastly expand.

“Orbital data centers obviously would be a complete game changer,” Martin said. “It solves three problems: the real estate problem, the power problem and the cooling problem.”

Ben Wild, chief technology officer of Hubble Network, a connectivity company that uses satellites to build a global bluetooth-based communications network for trucking and other industries, said he was once skeptical, but now believes space-based data centers are technically feasible and can make economic sense if certain conditions are met.

“From a technical standpoint, the fundamentals of orbital data centers are sound,” he said. “In the right orbit, you can generate continuous solar power and reject heat directly into space, removing two major constraints on terrestrial AI infrastructure.”

Wild said data centers in space could help address congestion at home, especially “as permitting delays, power constraints, and AI demand make new terrestrial capacity harder to build.”

SpaceX and Musk did not respond to a request for comment.

SpaceX laid out its case for putting data centers in space in documents ahead of its record-breaking initial public offering and its debut on the public markets last week, which resulted in Musk becoming the first trillionaire.

“[We] believe SpaceX’s reusable rockets, scaled satellite manufacturing, and operational expertise can enable the cost-effective and rapid deployment of massive AI compute satellite constellations — with potentially millions of satellites — for orbital data centers,” SpaceX said. “We believe these AI compute satellites in Sun-synchronous orbit will be able to handle energy-intensive AI workloads.”

But the company acknowledged the significant risks.

“Space is inherently hostile,” it said. “In particular, we have not, and no one else has, previously operated or attempted to operate orbital AI compute.”

Despite its lack of measurable progress on its two most ambitious goals, and its track record of bleeding money, SpaceX attracted massive investor interest upon debuting last week.

Its record-breaking $75 billion IPO preceded a rally that vaulted it into the orbit of tech unicorns. Musk threw fuel on the fire, predicting on Sunday that SpaceX “might be able to reach approximately $1T revenue in 2030,” before predicting revenue would probably exceed that milestone the following year.

For SpaceX to justify being worth as much as Apple or Microsoft, it would have to transform into the company its IPO documents suggest it can one day become. Already, it has its sights on becoming an artificial intelligence powerhouse. This week, it announced it had acquired the company behind the AI tool Cursor in a transaction valuing the artificial intelligence startup at $60 billion.

“The details of actually getting a million satellites into space and managing, monitoring [them] … seem pretty nontrivial to me. … There’s a lot of question marks around that,” said Martin. “They certainly are ascribing a ton of market cap to things that they have not yet demonstrated.”

In light of SpaceX’s gargantuan valuation, James Dow, professor of finance at London Business School, found himself asking questions about the nature of investors’ excitement.

“Are the people who are very enthusiastic about this — are they really knowledgeable, or is it more ideology and enthusiasm?”

“There is an ideological component, obviously,” he said, “because Musk presents himself that way.”