• Elon Musk estimates Terafab's AI compute output will be split roughly 25% for Tesla's Optimus robots and 75% for AI spacecraft.
  • The joint venture among Tesla, SpaceX, and xAI targets multi-hundred-billion-dollar scale with about 1 terawatt of AI compute per year.
  • The project faces significant capital requirements ($25B–$55B+) and execution risks, but could reshape AI hardware and robotics.

Musk's Vision for Terafab

Elon Musk provided a rare breakdown of the Terafab project's intended output, saying "my very approximate guess" is that AI compute would be about 25% for Tesla's Optimus robots and 75% for AI spacecraft. The comments came during a recent update on the ambitious venture, which aims to build a massive integration hub for semiconductors and AI compute, both in space and on Earth.

Terafab, a joint venture among Tesla, SpaceX, and xAI, was publicly announced around March 2026. The project plans to deploy a megafab in the Dallas/Austin, Texas area, with a factory footprint exceeding large automotive scales. The goal is to commoditize AI compute and semiconductor manufacturing, targeting roughly 1 terawatt of AI compute per year—a scale that would dwarf current capabilities.

The split reflects Musk's strategic priorities. Optimus, Tesla's humanoid robot, is expected to require substantial AI compute for real-time decision-making and mobility. Meanwhile, AI spacecraft—presumably for SpaceX's Starlink and interplanetary missions—would demand the majority, given the need for autonomous navigation, data processing, and communication systems.

Capital and Feasibility Concerns

The project's capital appetite is massive, with reports estimating costs between $25 billion and $55 billion or more. Such funding would rank among the largest industrial investments ever, rivaling major infrastructure projects. However, the timeline and return on investment remain uncertain, especially given global semiconductor cycles and geopolitical dependencies on chip supply chains.

"This is an incredibly bold bet," said one industry analyst who asked not to be named, citing the sensitivity of the matter. "The scale is unprecedented, but the execution risk is equally unprecedented."

Public reaction has been mixed, with some applauding the ambition while others question whether the numbers are realistic. Musk's own caveat—"very approximate"—suggests even the architects are unsure of the final output distribution.

Strategic Implications

If realized, Terafab could accelerate robotics, space-based AI, and centralized AI infrastructure. For Tesla, Optimus could become a major revenue driver, with Musk having previously stated that robots could eventually outnumber humans. For SpaceX, AI spacecraft would enable more autonomous missions and potentially support a self-sustaining Martian colony.

The venture also has implications for the broader AI industry. By producing compute at scale, Terafab could disrupt chip supply chains and AI hardware pricing, making advanced AI more accessible. However, that disruption comes with risks: capital intensity, potential delays, and reliance on advanced manufacturing techniques that are still evolving.

Reaction and Next Steps

Representatives for Tesla, SpaceX, and xAI did not respond to requests for comment. Musk's remarks were made in a context where he often shares rough estimates, but the specifics here offer a rare glimpse into the venture's planned output.

The Terafab project remains in early stages, with official details still evolving. Observers will be watching for updates on funding, site development, and technology milestones. As Musk himself acknowledged, the numbers are guesses—but they signal the scale of ambition behind one of the most talked-about industrial projects of the decade.

This article was updated to reflect Musk's remarks as "approximate" and to clarify the output split.