Cowboy Space asks FCC to approve 20,000-satellite Stampede after Google test craft reached orbit

The newest applicant for a data center in orbit is Cowboy Space Corporation, which has asked the Federal Communications Commission for permission to operate 20,000 satellites under the name Stampede. In the company's filing, chief operating officer Joseph Yaffe argues that Stampede "can skip the terrestrial power grid entirely and bypass the costs and delays associated with building data centers on Earth," and compresses the pitch into a phrase about putting "the silicon next to the sunlight." The commission has been receiving such claims with some regularity.

Cowboy Space is the fourth entrant to the category in about a year, and not the one asking for the most. A January filing from Elon Musk's SpaceX declared that "a constellation of a million satellites that operate as orbital data centers is the first step towards becoming a Kardashev II-level civilization." Blue Origin, the rocket company backed by Jeff Bezos, followed in March with Project Sunrise. Google moved first, unveiling Project Suncatcher in November 2025, when chief executive Sundar Pichai claimed the company would place data centers in space as early as 2027; the first test craft from that program reached orbit last month.

Data Center Dynamics, which assembled the sequence, traces its momentum to a stage rather than a laboratory. At Italian Tech Week in Turin last October, Bezos told a fireside audience that data centers "will be better built in space, because we have solar power there, 24/7," and over the months that followed the idea found favor with the investment community and with the tech industry's founders. Musk took it up as an ambition of the merged SpaceX and xAI.

The filings also mark a shift in what the largest technology fortunes are aimed at. DCD observes that Silicon Valley has set aside Martian colonies, asteroid mining and space tourism in favor of compute in orbit, and notes that firms had been eyeing orbital compute for years before this volume of paper reached the regulator. Three FCC applications and one corporate program have arrived in roughly a year.

DCD builds its analysis around a question the filings leave open: orbital data centers, or distributed LEO compute? The two would carry different owners, different counterparties and different return profiles, and nothing in the applications establishes which of Stampede, Project Sunrise or Suncatcher is meant to become which. For anyone underwriting the theme, that ambiguity is most of the risk.

What can be verified is narrower. Google's test craft is in orbit; DCD describes the other three as proposals and applications. The account names no customer for the capacity any of them proposes.

DCD's sharper reading is about purpose. Its suggestion is that strategic priorities may be cementing demand, that the applications matter less as descriptions of what exists today than as claims on what exists if orbital compute becomes real. On that view the commission is being asked to underwrite an option, and the applications read accordingly: they reach for civilizational scale well before they reach for a customer.

The terrestrial comparison is the one this desk keeps to hand. On the ground the scarce asset is power that already has a path to it: Blackstart Digital's agreement to buy IBM's Almaden campus and the 25MW substation attached to it, and Orbital, a 10-gigawatt constellation with $5 million in the bank and a filing still pending. In orbit, the claim is a power source that requires no substation and no interconnection queue. The pitch in both places is relief from the same constraint; the difference is that a substation can be inspected, and an orbital power advantage can, so far, only be argued.

That leaves the practical question for an allocator: what would an order from the commission actually authorize? A grant would give an applicant the right to fly satellites, which is not a cost per unit of compute, and nothing in the record prices the one against the other. Cowboy Space's application is the fourth entrant in a year. Whether it, SpaceX, Blue Origin or Google is building an orbital data center business or a distributed compute network in low earth orbit is the question the first order out of the queue will have to answer.