Starcloud

The co founder of @Starcloud_ , @PhilipJohnston , just talked about the single most important number in the space sector The most important number in the space sector right now is $500 per kilogram Starcloud’s CEO just laid out why, and it reframes orbital compute from science fiction into a commodity arbitrage with a crossing point Think of a data center as a solar project. On Earth, terrestrial solar has three dominant costs: permitted land, batteries for when the sun isn’t up, and the cells themselves Move it to dawn-dusk sun-synchronous orbit and three things happen: 1- Permitted land goes to zero. Your single largest cost line disappears, along with the interconnect queue, the zoning fight, and the local referendum 2- Batteries go to zero. That orbit is in sunlight 24/7. No storage, no backup generation, no duck curve 3- You need 8x fewer cells. A square meter of panel in orbit produces roughly eight times what it produces on the ground. No atmosphere, no night, no weather. Three of your four cost lines collapse. One new line appears: launch So the entire thesis reduces to a single variable. Below roughly $500/kg to orbit, space undercuts land. Falcon 9 is about $5,000/kg today. Starship’s internal design target is $20–40/kg Why this is structurally bullish for the sector Because it converts launch from a cost center into the toll booth on a new asset class. Every dollar of launch price reduction doesn’t just make existing missions cheaper, it unlocks a category of demand that was economically impossible the day before Orbital compute is the first demand curve in space history that scales with AI capex rather than with government budgets, and AI capex is currently the largest private infrastructure buildout on Earth It also inverts the scarcity: Terrestrial data centers are now constrained by power and permitted land, not chips. Orbit has effectively unlimited uncontested power and no permitting authority. Starcloud claims ~10 TW fits in that one orbital band, about 20x the entire US grid Where the value actually accrues then? Launch is the choke point and the toll booth. Everything downstream is a derivative of $/kg. Watch Starship upper-stage catch, then re-flight, then, critically, whether SpaceX sells commercial capacity at all from 2028. On Falcon 9 they charge 2–3x marginal cost. If they hold that discipline on Starship, the whole sector clears. If they keep capacity in-house, they become the competitor, not the supplier Thermal is one of starcloud's real moat, and it’s underpriced. Everyone fixates on radiation, the operators say cooling was the hard problem and radiators are where the engineering differentiation will make it or break it imo