Elon Musk confirmed on Saturday that SpaceX has been quietly building a foundry in Bastrop, Texas to cast the blades and vanes inside natural gas turbines — a component so difficult to manufacture that only four companies worldwide can produce them at industrial scale. The move is a direct response to a constraint now rivaling GPU shortages as AI’s biggest headache: the physical power grid.
The bottleneck is real. GE Vernova says it’s essentially sold out of turbine production capacity through 2030, largely because hyperscalers like Amazon, Google, Meta, and Microsoft have all pivoted to building private gas-fired plants next to their data centers rather than waiting years for grid interconnections. The IEA projects global data center electricity demand will roughly double by 2030. Musk’s claim is that in-house casting at SpaceX could accelerate turbine delivery by up to 18 months. “The limiting factor for nat gas turbine production is casting the blades & vanes,” he wrote on X.
What makes the casting so brutal is the metallurgy. The blades operate at 3,000 to 3,600 degrees Fahrenheit — roughly 800 degrees hotter than the melting point of the alloy they’re made from. They survive only because each blade is cast as a single, unbroken crystal, grown slowly inside a vacuum furnace without the microscopic seams that let ordinary cast metal crack under stress. That’s challenging enough for jet engine blades; power-plant turbine blades are considerably larger, making defect-free production even harder.
If SpaceX actually pulls this off, a Musk-controlled entity would hold a manufacturing capability that every other AI infrastructure builder currently depends on a tiny oligopoly for. That’s a structural advantage no amount of venture funding can replicate quickly. But the strategy carries real baggage. In Memphis, where xAI has run gas turbines to power its Colossus data center since 2024, the NAACP has repeatedly accused the company of operating without required permits or pollution controls. A Virginia study using the EPA’s own COBRA model estimated that just eight full-time turbines at one facility could cause 3.4 to 6.5 additional premature deaths annually, translating to $53 million to $99 million in health-related damages each year. Musk is solving one bottleneck; he may be deepening another.
