There aren’t enough rockets for space data centers. Cowboy Space raised $275 million to build them.

Cowboy Space Corporation raised $275 million in a Series B at a $2 billion post-money valuation, led by Index Ventures, to fund development of its own rocket program for orbital data centers. The startup says it expects its first launch before the end of 2028 and plans to build 20,000-25,000 kg satellites that generate 1 MW of power for nearly 800 GPUs each. The article highlights strong AI compute demand, but also major execution risk and limited launch capacity across the commercial space industry.

Analysis

The real market signal is not “space data centers” per se; it is that launch capacity has become the gating input for every orbital compute thesis. That creates a near-term bottleneck that advantages the incumbent launch stack and any supplier with scarce propulsion, range, or regulatory know-how, while pushing most orbital-AI concepts into a long-dated venture option rather than a scalable infrastructure market. The second-order effect is that capital may increasingly migrate from application-layer space ideas to the picks-and-shovels layer: engines, avionics, test infrastructure, and launch ops.

For public-market names, the risk/reward is asymmetric but not uniform. Blue-sky space narratives often get priced as if launch cadence is the only constraint; in reality, the bigger issue is that vertically integrated launch providers can ration capacity to their own payload roadmaps, which delays third-party TAM realization by years. That hurts any company that depends on an external manifest and favors providers that control both rocket and payload economics, even if commercialization slips.

The contrarian view is that this is less an immediate revenue opportunity than a financing and credibility test. A $2B valuation can support bold language, but the path from prototype to reusable orbital launch is dominated by capex, testing failures, and regulatory delay; the implied timeline suggests dilution risk well before any meaningful revenue. In the interim, the market may overestimate how quickly AI demand translates into orbital deployment and underestimate how much terrestrial solutions can still improve cost/performance over the next 24 months.

Key Decisions for Investors

• Avoid chasing the theme in near-term commercial launch beneficiaries; treat orbital-data-center names as venture-like options with a 3-5 year realization window, not a 6-12 month catalyst.
• Keep a tactical long bias in scarce launch-enablement suppliers and private-market exposure where possible; for public proxies, favor names with propulsion or range infrastructure exposure over pure launch aspirants.
• Use weakness in FLY as a relative-value short against more capable launch incumbents only if the company’s funding runway and test cadence deteriorate; otherwise the better trade is to avoid owning development-stage launch at this point in the cycle.
• If you want thematic exposure, structure it as a call-spread on broader aerospace/launch infrastructure rather than outright equity: the upside comes from duration extension of the launch shortage, while downside is capped if orbital compute remains a niche.
• Monitor SAIC for contract spillover from space domain awareness and launch/test infrastructure demand, but do not assume direct earnings uplift; this is a watchlist name, not an immediate catalyst trade.