SpaceX is moving deeper into energy infrastructure as Elon Musk targets the growing power demands of artificial intelligence.
SpaceX is moving deeper into the infrastructure behind artificial intelligence, and this time the target isn’t chips, rockets or satellites. It’s electricity.
Elon Musk has confirmed that SpaceX plans to manufacture critical components for natural-gas turbines in-house, specifically the blades and vanes whose limited supply has become a bottleneck for new power generation.
Musk says bringing the casting process inside SpaceX could allow natural-gas turbines to come online as much as 18 months sooner.
It sounds like an obscure manufacturing story.
It isn’t.
This is increasingly becoming one of the biggest stories in artificial intelligence because the AI race is quickly turning into an energy race.
AI Has a New Problem: Finding Enough Electricity
For the past several years, the AI conversation centered primarily around computing power.
Companies raced to secure GPUs, construct enormous data centers and train increasingly sophisticated models.
But having millions of advanced chips doesn’t matter much if there isn’t enough electricity to turn them on.
The International Energy Agency expects global data-center electricity consumption to roughly double by 2030. Meanwhile, GE Vernova’s gas-turbine production capacity is effectively committed through 2030 as utilities and data-center developers compete for additional generating capacity.
That has created another infrastructure race underneath the AI boom.
Technology companies need power plants.
They need substations.
They need transmission.
They need natural gas.
They need solar generation and battery storage.
And increasingly, they need gas turbines.
Companies including Amazon, Google, Meta, Microsoft and OpenAI have explored or deployed natural-gas solutions as the industry searches for ways to bring enormous amounts of electricity online faster.
Musk now wants to attack one of the industry’s most stubborn manufacturing constraints.
Why Turbine Blades Suddenly Matter to Artificial Intelligence
Industrial gas turbines operate under extraordinary temperatures and stresses.
That makes their internal blades and vanes incredibly difficult to manufacture.
These aren’t ordinary pieces of metal stamped out on an assembly line. High-performance turbine components can involve sophisticated nickel-based superalloys, precision casting and extremely demanding manufacturing tolerances.
Production can also take an extraordinary amount of time.
Reports indicate turbine-blade manufacturing can require 60 to 90 weeks from beginning to completion.
That creates a problem.
AI infrastructure can potentially be constructed faster than the power equipment needed to operate it.
Musk believes SpaceX can shorten that timeline.
“The limiting factor for nat gas turbine production is casting the blades & vanes,” Musk wrote in a recent post, adding that SpaceX’s in-house casting could accelerate turbines coming online by as much as 18 months.
That is the real significance of the announcement.
SpaceX isn’t simply trying to manufacture another component.
It is trying to remove a bottleneck from the AI supply chain.
SpaceX Is Reportedly Building the Operation in Texas
The manufacturing effort is taking shape around Bastrop, Texas, where SpaceX already has a significant presence.
The Information reported that SpaceX has been laying the groundwork for a specialized foundry to manufacture turbine blades and vanes. Job listings and other indicators pointed toward the project before Musk publicly confirmed the strategy.
The location also fits Musk’s broader expansion across Texas.
But more importantly, SpaceX already possesses something few technology companies have: deep expertise in highly specialized manufacturing.
SpaceX became successful partly because it rejected the traditional aerospace model of outsourcing enormous portions of production.
Instead, the company vertically integrated much of its manufacturing.
Now Musk appears ready to apply a similar philosophy to energy infrastructure.
If a supplier can’t manufacture something quickly enough, SpaceX may simply manufacture it itself.
The SpaceX Strategy Is Bigger Than Space
This move illustrates how dramatically SpaceX’s role is evolving.
The company built its reputation around rockets.
Then came Starlink.
Now SpaceX’s broader ecosystem intersects with communications, AI infrastructure, energy and advanced manufacturing.
That vertical-integration strategy has long been central to Musk’s companies.
When a critical component becomes expensive, scarce or slow to arrive, Musk’s instinct frequently appears to be the same:
Bring production in-house.
For AI, the logic becomes particularly powerful.
Every month that a billion-dollar data center sits waiting for electricity represents enormous amounts of computing capacity that cannot generate revenue or train models.
Saving 18 months on power infrastructure could therefore have implications worth far more than the cost of the turbine components themselves.
Natural Gas Is Becoming AI’s Bridge Fuel
Musk is simultaneously pushing solar production.
He said SpaceX and Tesla are each working toward enormous solar manufacturing capacity. However, he also acknowledged that natural gas will remain necessary for several years to supplement and help bootstrap solar generation.
That highlights an uncomfortable reality surrounding the AI boom.
Artificial intelligence may ultimately accelerate advances in energy efficiency and clean technology, but today’s enormous AI clusters require staggering amounts of reliable electricity right now.
The existing electrical grid cannot always provide it quickly enough.
Gas turbines offer something hyperscale AI developers desperately want: power that can potentially be installed directly near a data center without waiting years for major grid upgrades.
SpaceX has already made enormous commitments to the technology. Earlier this year, reporting based on regulatory filings showed SpaceX had committed more than $2.8 billion toward gas turbines for AI data centers.
That makes the new manufacturing strategy considerably more consequential.
Musk isn’t experimenting with turbines.
He is building an energy strategy around them.
There Is an Environmental Cost
There is another side to the story.
Natural-gas turbines generate carbon emissions and other pollutants, and Musk’s use of turbines for AI infrastructure has already faced scrutiny.
SpaceX’s turbine deployments connected with xAI’s Colossus facilities have generated community complaints, regulatory questions and environmental concerns.
That tension will become increasingly important as AI infrastructure expands.
The industry wants enormous amounts of electricity.
Communities want economic investment.
Governments want leadership in artificial intelligence.
But rapidly deploying fossil-fuel generation alongside data centers could create new battles over emissions, permitting, water consumption and local infrastructure.
The AI revolution may therefore create an entirely new debate over America’s energy policy.
The AI Race Is Becoming an Industrial Race
Perhaps the biggest takeaway from SpaceX’s turbine project is how much the artificial-intelligence competition has changed.
AI is no longer simply about who has the best model.
The winners increasingly need control over an enormous physical supply chain:
Chips → servers → data centers → cooling → transmission → electricity → power generation.
And now, apparently, turbine blades.
That makes the companies capable of combining software with manufacturing, energy and infrastructure increasingly powerful.
SpaceX’s move also puts traditional turbine-component manufacturers under a new spotlight. Shares of Howmet Aerospace dropped sharply following Musk’s comments, although several Wall Street analysts argued that the reaction may have been excessive given the difficulty of entering this highly specialized industry.
Manufacturing these components at scale will not be easy.
The Wall Street Journal reports that turbine components can require advanced single-crystal nickel-superalloy manufacturing, specialized vacuum furnaces and highly skilled technicians. Industry experts caution that developing reliable production capacity could take years.
So Musk’s 18-month acceleration target should be viewed as an ambition rather than a guaranteed outcome.
The Bottom Line
For years, Silicon Valley’s AI race looked like a competition for algorithms and GPUs.
Now the industry’s biggest constraints are becoming increasingly physical.
Electricity.
Land.
Cooling.
Transmission.
Transformers.
Turbines.
And even the individual blades inside those turbines.
SpaceX’s decision to manufacture turbine blades and vanes is another sign that the next phase of artificial intelligence will not be won entirely inside a computer lab.
It will also be won in factories, foundries, power plants and data centers.
Elon Musk appears to understand that.
And if SpaceX can genuinely remove 18 months from the timeline required to deploy new power generation, it won’t simply be solving a turbine shortage.
It could be giving Musk’s AI ambitions something increasingly valuable in the race for artificial intelligence: more power, faster.
About the Author
