Why Is SpaceX Priced at $2.5 Trillion? Computers in Space?
$2.5 trillion is the price tag MarketWatch gives SpaceX, based on a plan for AI data centers in orbit. The outlet calls that trillion-dollar valuation a bet on the orbital compute dream. Its Sept. 28 reporting describes a massive AI data center in orbit that would use sunshine and the vacuum of space to solve power and cooling problems. Think of it like a factory in space running on endless sunshine. MarketWatch
SpaceX talks about it in similar words. It says it wants to capture solar energy in space to run low-cost, high-performance AI computers for Earth. Its AI satellite page says each payload, the computer load carried in space, can use up to 250 kW peak / 175 kW average. A kW is a unit of electric power. SpaceX Power is the bottleneck. Cooling is the second.
The company also says today's AI progress depends on big data centers on Earth that need huge amounts of power and cooling. Its IPO roadshow presentation from June 11 calls its AI computer system terrestrial and orbital, meaning on Earth and in space. The same presentation calls solar power in space unlimited, clean, and lower cost. SpaceX IPO Roadshow
The Earth part is already described. SpaceX says its supercomputing buildings in Memphis, TN / Southaven, MS rank among the world's most advanced AI training clusters, groups of computers joined to teach AI. It says those Mid-South sites cover over 2.5 million feet and hold millions of GPUs, the special chips that do AI work. SpaceX
Schedules and scale remain proposals. Reuters reported on June 9, 2026, citing sources, that SpaceX aims to launch test flights for orbital AI computing by the end of next year. Reuters Aerospace America reported on July 1 that SpaceX and others have proposed launching tens of thousands of massive satellites for AI processing and other functions. Aerospace America Bloomberg reported on May 14 that SpaceX and Blue Origin both announced plans to build and launch orbital data centers. Bloomberg A former Meta chief technology officer said SpaceX is the only company capable of building orbital data centers.
The broader context here is how the $2.5 trillion is split. Part pays for computers that exist now. Part pays for a chance at cheap power later. The Earth buildings give experts something solid to measure now: building size, chip count, and a claim to top-level AI training in the Mid-South. The space part is different. It depends on future launch costs, delivering power in orbit, and using a 250 kW peak / 175 kW average payload at large scale.
In my view, that split explains why the roadshow says terrestrial and orbital. Earth computers can bring training income soon. Space computers are a long-term bet on cheaper energy. If sunshine in space is treated as unlimited, clean, and lower cost, the hard part moves from buying electricity and cooling on Earth to building enough satellites, launching them often, and paying for tens of thousands of units. That takes a lot of upfront cash for hardware. Markets will judge it on cost per usable kilowatt, how fully the machines are used, and what interest rate is used to value future cash from tests that have not flown yet.
Looking at what this means for valuation discipline, the $2.5 trillion puts the risk on execution, not demand. Demand for AI computers is the stated reason. Lack of cheap power and cooling is the stated limit. Tests planned for the end of next year are the next real check. Until then, the Earth cluster is the anchor, and orbit is expectation.
