SpaceX’s first public earnings reveal a company whose rockets still define its identity, but whose financial future is increasingly tied to satellite connectivity and an enormously capital-intensive bet on selling AI compute. (Source: Image by RR)

Musk’s Biggest Investments Increasingly Happen on the Ground

SpaceX’s first quarterly earnings report as a public company paints a picture that differs considerably from the rocket-focused identity associated with Elon Musk’s most celebrated business. According to an article in theverge.com, the company’s launch operation generated less than $1 billion during the quarter. It accounted for only slightly more than 10 percent of total revenue, with SpaceX itself remaining one of its largest launch customers. By contrast, Starlink generated approximately $4.2 billion and was the only major segment to report an operating profit, making connectivity—not rockets—the company’s strongest established business.

The company’s most aggressive investment, however, is now occurring in artificial intelligence infrastructure. SpaceX reportedly spent $15.8 billion on AI during the second quarter, compared with slightly more than $1 billion each on its space and connectivity businesses. Much of that spending supports a rapidly expanding “neocloud” operation that rents computing capacity to outside AI companies. What began as infrastructure built largely for xAI’s Grok has increasingly become a commercial compute business serving customers including Google and Anthropic after xAI encountered difficulties efficiently utilizing the infrastructure itself.

SpaceX executives are projecting extraordinary growth from this new business. The company says contracts and other contributions could put it on a path toward a $100 billion annualized revenue run rate, while Musk continues promoting even more ambitious plans involving orbital data centers, a massive Terafab semiconductor operation, fleets of AI satellites, Optimus robots, and eventually a lunar mass accelerator. The article treats those longer-range claims skeptically, arguing that SpaceX must first demonstrate that its expensive terrestrial data-center operation can become sustainably profitable before investors assign substantial value to even more speculative projects.

The broader argument is that today’s SpaceX increasingly resembles a vertically integrated telecommunications and computing conglomerate that happens to own the world’s most prominent private rocket business. Starlink provides its strongest operating business, compute represents its largest growth bet, and launches increasingly support SpaceX’s own infrastructure. With insider lockups beginning to expire and shares already under pressure, the author argues that investors should distinguish between SpaceX’s extraordinary space ambitions and the considerably more terrestrial businesses currently responsible for most of its economics.

read more at theverge.com