SpaceX has quietly transformed from a rocket company into a data center operator, with artificial intelligence infrastructure now generating more revenue and consuming far more spending than its original space business. According to the company's first quarterly earnings statement as a public company, the space sector contributed just over 10 percent of total revenue and didn't break $1 billion for the quarter, while data center leasing and satellite internet services now dominate the business model.

What Happened to SpaceX's Core Space Business?

The numbers tell a striking story about where Musk's priorities have shifted. SpaceX remains its own biggest customer for rocket launches, and there simply aren't enough other customers willing to pay for the service. Meanwhile, the company's spending on AI infrastructure reached $15.8 billion in the second quarter alone, compared to just over $1 billion combined for space and connectivity sectors.

This dramatic reallocation of resources stems from challenges with Grok, Musk's artificial intelligence system. The company built Colossus 1, a massive data center in Memphis specifically for training Grok, but encountered technical problems including latency issues that made in-house model training difficult. The facility also suffered from a mix of newer and older computer chips that created bottlenecks. Rather than fix these problems, SpaceX decided to lease the data center capacity to other AI companies instead.

How Is SpaceX Making Money From AI Infrastructure?

SpaceX has signed major deals with Google, Anthropic, Reflection AI, and Cursor, an AI company that Musk eventually chose to acquire. These partnerships have put the company on what executives describe as a path to reach $100 billion in annualized revenue run rate, a financial measurement that estimates yearly income based on shorter-term performance. On the earnings call, SpaceX's chief financial officer stated the company is "on a trajectory, including contribution from Cursor, to reach $100 billion of ARR," while Musk declared that "the $100 billion ARR in December is not a question mark".

financial officer

The satellite internet business, Starlink, remains the only profitable division. It generated $4.2 billion in revenue and was the only part of SpaceX that didn't operate at a loss. However, even Starlink's future is being reshaped around AI ambitions, with plans announced for a phone service to compete with AT&T, Verizon, and T-Mobile.

What Are the Risks of SpaceX's Data Center Strategy?

• Commodity Competition: The bare-metal compute business operates on razor-thin margins because computing power is essentially a commodity. As more data centers get built globally, the available compute increases, forcing companies to compete primarily on cost and making it harder to maintain premium pricing.
• Capital Intensity: Building data centers requires enormous upfront investment in physical infrastructure, power systems, and cooling equipment. Revenue doesn't equal profit, and the expenses of construction and maintenance can quickly erode earnings, especially if demand softens.
• Technological Obsolescence: Computer chips become outdated relatively quickly as newer, more powerful processors emerge. Data center operators must constantly upgrade equipment or risk offering customers inferior performance compared to competitors using newer technology.

Despite these challenges, Musk has announced even more ambitious plans. He claims he took SpaceX public specifically to build data centers in space, proposing an orbital data center consisting of as many as 1 million satellites to the Federal Communications Commission. The application is notably light on technical details, raising questions about whether it's a serious proposal or primarily a public relations move.

What Is Musk's Vision for Space-Based Computing?

Musk's long-term vision involves several interconnected pieces. He plans for a chip producer called Terafab to manufacture one terawatt of chips annually. A billion Optimus robots, once Musk figures out how to build functional hands, would theoretically do the work. The ultimate goal, according to Musk, is to build a mass accelerator on the Moon.

These proposals sound expensive because they are extraordinarily expensive. Critics note that if Musk struggled to run a data center on the ground, attempting something far more difficult in space seems premature. However, Musk's track record suggests he often pursues ambitious goals despite technical challenges. The question remains whether these space-based computing plans will ever move beyond concept drawings and earnings call rhetoric.

What's clear from SpaceX's first earnings report is that the company Musk founded to reach Mars has become something quite different: a terrestrial infrastructure company with space ambitions on the side. The rockets that once defined SpaceX are now secondary to the data centers that actually generate revenue and consume the company's resources.

" }