SpaceX Orbital Data Centres Could Cost Five Trillion Dollars a Year, Analysts Warn

The Biggest Merger in History

SpaceX has completed its acquisition of Elon Musk's artificial intelligence startup xAI in a deal valued at one point two five trillion dollars, making it the largest corporate merger of all time. The transaction values SpaceX at one trillion dollars and xAI at two hundred and fifty billion dollars, surpassing the previous record set by Vodafone's acquisition of Mannesmann in two thousand.

A Million Satellites in Orbit

Alongside the merger, SpaceX has filed with the Federal Communications Commission seeking authorisation to launch up to one million satellites to function as orbital data centres. The filing describes the system as a high-bandwidth, optically linked constellation of solar-powered satellites with unprecedented computing capacity to power advanced AI models. Musk has claimed that within two to three years, the most cost-effective method to generate AI computing power will be in space.

The Five Trillion Dollar Question

Analysts at MoffettNathanson have published a sobering assessment of the economics. The research firm estimates that a full-scale buildout could require between two trillion and five trillion dollars annually in capital expenditure, potentially one-sixth of America's gross domestic product. Maintaining a million-satellite constellation would require launching roughly two hundred thousand satellites per year, assuming each lasts five years, translating to approximately three thousand three hundred Starship launches annually, or about nine per day.

Engineering Challenges Mount

Beyond cost, experts highlight significant technical hurdles. In the vacuum of space, there is no air to cool processors, requiring massive radiator panels. A single high-end GPU would need roughly one square metre of radiator surface. Modern semiconductors are not designed to endure space radiation, which could impair reliability. Deutsche Bank analysts predict the first small-scale orbital data centre trials will not occur until twenty twenty-seven or twenty twenty-eight.

IPO Narrative or Genuine Strategy?

AWS chief executive Matt Garman has dismissed orbital data centres as pretty far from reality, noting that the cost of getting payloads into space remains massive. Several analysts suggest the orbital data centre pitch is strategic positioning for SpaceX's anticipated mid-twenty twenty-six IPO, which could raise up to fifty billion dollars, rather than an imminent business line. Meanwhile, xAI continues burning roughly one billion dollars per month, suggesting the merger may be as much about financial lifelines as technological ambition.