Affiliate disclosure: Some links in this article may be affiliate links. We may earn a commission if you purchase via these links — at no extra cost to you. This does not affect our editorial coverage. Full disclosure.
Robinhood Co-Founder Raises $275M to Put AI Data Centres in Orbit — 800 GPUs Per Satellite
TL;DR
Baiju Bhatt, co-founder of Robinhood, raised $275 million at a $2 billion valuation for Cowboy Space Corporation, which plans to build rockets whose upper stages double as AI compute nodes in low Earth orbit. Each satellite carries 800 GPUs powered by solar energy. The company's constellation is called Stampede and aims to launch before the end of 2028.
$275M
Series B raise at $2B valuation, led by Index Ventures
800
GPUs per satellite, solar-powered in low Earth orbit
2028
target year for first Stampede constellation deployment
Cowboy Space Corporation announced a $275 million Series B on May 11, led by Index Ventures with participation from IVP, Blossom Capital, SAIC, Andreessen Horowitz, Breakthrough Energy Ventures, and others. The company — formerly operating as Aetherflux — was founded in 2024 by Baiju Bhatt, who built Robinhood into a $40B+ brokerage. The valuation at close: $2 billion.
The concept is unconventional. Cowboy Space is not buying compute time in an existing data centre, nor building one on the ground. It is designing a rocket whose second stage becomes a data centre after delivering its payload — staying in low Earth orbit with 800 onboard GPUs, a 20,000–25,000 kg mass, and 1 MW of solar-generated power per satellite. The orbital constellation is named Stampede.
Why orbit? The AI compute shortage on Earth is severe — land, power, water, and cooling are all constrained. Power grids in major data centre markets (Virginia, Texas, Singapore) are at or near capacity. The argument for orbit: unlimited solar power with no cooling costs, no land permitting, no power grid competition. The counter-argument: latency, launch cost, and the fact that no one has done this before at meaningful scale.
TechCrunch's framing is pointed: *'There aren't enough rockets for space data centres — Cowboy Space raised $275M to build them.'* Cowboy Space is not just launching compute; it is building its own rocket to launch that compute, because existing launch capacity is insufficient for the number of orbital data centre companies now competing for slots. The capital required to be vertically integrated — rocket manufacturer plus orbital data centre operator — is why the raise is $275M at seed-to-Series-B stage.
The timeline is ambitious. The company targets its first constellation deployment before the end of 2028. That gives it roughly 30 months to complete rocket development, satellite manufacturing, regulatory approval, and initial launches — a timeline that would be aggressive for a mature launch provider, let alone a 2024 startup.
Why It Matters
The AI compute shortage is now so acute it is pushing investment into space. That sentence would have read as science fiction two years ago. The signal here is not that Cowboy Space will definitely succeed — orbital data centres remain unproven at scale and the 2028 timeline is aggressive. The signal is that serious investors with serious capital are backing the premise that ground-based AI infrastructure cannot keep up with demand. For anyone building AI-powered products and services: the platforms you build on top of are capacity-constrained, and that constraint is driving the entire infrastructure layer to explore solutions that would have been dismissed as impractical a decade ago. The demand side of AI is not slowing down. The supply side is racing to catch up by any means available — including orbit.
Who's Affected
- — AI infrastructure buyers (hyperscalers, large enterprises) — the orbital data centre market, if it delivers, offers a new supply source for GPU compute currently in extreme shortage
- — Existing launch providers (SpaceX, Rocket Lab, Blue Origin) — Cowboy Space entering the rocket manufacturing market adds competitive pressure and signals demand is outstripping capacity
- — AI product builders dependent on GPU availability — the medium-term supply picture for compute may improve faster than current ground-based construction timelines suggest
- — Climate-focused observers — solar-powered orbital compute eliminates the grid power and cooling water demands that make large ground-based AI data centres environmentally contentious
What To Do Now
- 1. If you are building AI products today, the compute shortage affecting your API pricing and rate limits has a multi-year resolution timeline. Budget for constrained supply through at least 2027 in your cost models.
- 2. If you follow AI infrastructure, track Cowboy Space alongside Crusoe Energy, CoreWeave, and Lambda Labs — the supply-side race to provide GPU compute is a meaningful indicator of where AI product economics are heading.
- 3. The orbital angle is speculative but not frivolous. Index Ventures, a16z, and Breakthrough Energy are not known for novelty investments — they backed this because the demand case is real and the technical barriers, while significant, are not physics-level impossible.
- 4. For independent builders: nothing changes this week. This is a 2028+ story. But it is a useful data point when someone asks whether AI compute costs will come down — the answer is yes, but the timeline is longer than most assume.
More on this topic — Best AI Tools 2026
The AI Hustle Playbook Newsletter
Get the curated shortlist.
A playbook of AI tools and strategies for building income streams.
No spam. Unsubscribe anytime.
More from ToolNav News
Runway Hits $5.3B Valuation After Adding $40M ARR in One Quarter — Now It Wants to Beat Google at AI Video
2026-05-17
Anthropic in Talks for $30–$50B Raise at Up to $950B Valuation — Would Surpass OpenAI for the First Time
2026-05-15
Cursor Lands in Microsoft Teams + Switches Bugbot to Usage-Based Billing — Solo Devs Get a Real Pricing Reset
2026-05-15