Search
TechMallowTech news · Reviewed properly

// SCIENCE

SpaceX Builds Foundry for Gas Turbine Parts to Power AI Data Centers

Elon Musk's SpaceX is constructing a specialized foundry in Bastrop, Texas, to manufacture gas turbine blades and vanes, a move intended to accelerate the deployment of natural gas turbines for AI data centers.

On this page
An industrial facility showing large machinery for casting turbine components.
TechMallow
TL;DR
  • SpaceX is building a foundry in Bastrop, Texas, to produce gas turbine blades and vanes.
  • This initiative aims to accelerate natural gas turbine deployment by up to 18 months.
  • The move addresses power grid constraints facing the artificial intelligence industry.
  • Manufacturing these turbine parts is highly complex due to extreme temperatures and material requirements.
  • Environmental concerns exist regarding increased gas turbine use and associated pollution.

Elon Musk's SpaceX is constructing a specialized foundry in Bastrop, Texas, to manufacture critical gas turbine components, a development Musk confirmed on Saturday. This facility is designed to produce blades and vanes for natural gas turbines, potentially accelerating their availability by up to 18 months. The initiative aims to address a significant bottleneck in providing power for the rapidly expanding artificial intelligence (AI) industry.

The existence of the secret foundry was first suggested by job listings and findings from due diligence specialist Corey Trinetti, who reported that [SpaceX] acquired approximately 830 acres near its existing Starlink factory in Bastrop between March and June. Elon Musk later stated on X that both [SpaceX] and Tesla are working to build 100 gigawatts per year of solar production capacity. However, he noted that natural gas would be necessary to supplement and bootstrap solar power for several years. Musk identified the casting of blades and vanes as the limiting factor for natural gas turbine production, calling in-house casting a profound game-changer.

Addressing AI's Power Demands

The push for faster turbine production stems from the immense power demands of the [AI] industry. While GPU shortages remain an issue, the physical power grid has emerged as another critical constraint. The International Energy Agency forecasts that global data center electricity use will roughly double by 2030. GE Vernova, a major gas turbine manufacturer, has reported that its production capacity is essentially sold out through 2030, largely due to the demand from [AI] infrastructure.

To circumvent these grid limitations, major hyperscalers such as Amazon, Google, Meta, OpenAI, and Microsoft are increasingly adopting a strategy of building private gas-fired power plants adjacent to their data centers. After years of prioritizing renewable energy sources like wind and solar, these companies are now relying on natural gas to bring new data centers online more quickly.

The Complexities of Turbine Blade Manufacturing

Manufacturing the blades for gas turbines presents a significant technical challenge. These blades operate within the hottest section of a turbine, enduring temperatures between 3,000 and 3,600 degrees Fahrenheit. This is approximately 800 degrees hotter than the melting point of the metal alloy from which they are made. Their survival at such extreme temperatures is attributed to intricate internal cooling channels, thermal-barrier coatings, and a highly specialized casting process.

Only four companies globally possess the expertise to produce these specific blades at an industrial scale, and all are currently operating at full capacity. The difficulty lies in casting each blade as a single, unbroken crystal. This process requires slow growth within a vacuum furnace to avoid microscopic seams that could cause ordinary cast metal to crack under stress. While challenging for smaller jet engine blades, the considerably larger blades used in power-plant turbines make large-scale, defect-free production even more arduous.

Environmental Impact and Regulatory Scrutiny

Should [SpaceX] successfully master this complex manufacturing process, a Musk-controlled entity would gain a critical capability that other [AI] infrastructure developers currently depend on a small group of specialized suppliers for. This could provide SpaceXAI with a competitive advantage that would be difficult for non-manufacturing competitors to replicate quickly.

However, the accelerated deployment of natural gas turbines raises environmental concerns. Gas turbines are known to emit pollutants, and their use has already led to federal lawsuits and peer-reviewed health research. In Memphis, where SpaceXAI has been operating gas turbines for its Colossus data centers since 2024, the NAACP has repeatedly accused the company of running turbines without the necessary permits or pollution controls mandated by federal law. The organization has expressed concern that these turbines release smog-forming compounds and hazardous chemicals like formaldehyde, which are linked to health issues such as asthma and respiratory diseases.

Was this article helpful?

More from Ines Kowalski