Elon Musk says Tesla will build and operate its planned "Terafab" itself, directly ruling out TSMC or any external foundry partner for the facility.

The statement closes off what had been an open question about how Tesla would scale manufacturing capacity for its compute silicon and AI ambitions. Rather than contracting a merchant foundry to stand up the enormous production footprint implied by the Terafab concept, Tesla intends to own the construction and the operation, according to Musk's remarks reported by TeslaNorth.com.

What did Musk actually commit to?

Musk's position is categorical: Tesla builds it, Tesla runs it. The name itself signals the scale the company has in mind — a facility sized in "tera" terms, echoing Tesla's existing terminology for high-throughput plants.

Notably absent from the statement is any delegation of the project to TSMC, the world's largest contract chipmaker and the foundry most observers assumed would anchor any Tesla move into serious silicon production at scale.

Why does going solo matter?

A self-built, self-run fab is a materially different commitment than a foundry partnership. Tesla would carry the full capital burden of construction, cleanroom fit-out, equipment procurement and process operations — costs that merchant foundries normally amortize across many customers.

It also means Tesla would need to recruit and retain the specialized workforce that fabs require, a scarce talent pool that established players including TSMC, Samsung and Intel already compete over aggressively.

The flip side is control. Owning the fab end to end would give Tesla full authority over capacity allocation, process priorities and scheduling — no queue behind other customers, no negotiating for wafer supply in a tight market.

What is confirmed versus what is not?

Confirmed, per Musk's statement as reported by TeslaNorth.com:

• Tesla intends to build the Terafab itself
• Tesla intends to run the Terafab itself
• TSMC is not part of the plan

Not yet established:

• A site location
• A construction timeline
• A capital budget
• Target process nodes or wafer sizes
• Which products the fab would produce

Musk's declaration is therefore a strategic direction, not a shovel-ready project. The gap between intent and ground-breaking remains wide, and Tesla has not published figures that would allow analysts to size the investment.

What does this mean for the foundry picture?

For TSMC, the immediate commercial impact is limited — Tesla has never been confirmed as a major merchant-foundry customer at AI-accelerator scale, and TSMC's order book is dominated by established hyperscaler and smartphone demand.

The longer-term signal matters more. If Tesla executes a vertically integrated fab strategy, it joins a short list of companies — Intel, Samsung, and the U.S. national-lab-adjacent programs among them — betting that owning leading-edge manufacturing is worth the capital and operational risk.

Musk's companies have form here. Tesla already vertically integrated battery, vehicle and, with xAI, datacenter infrastructure rather than renting capacity. The Terafab statement extends that same logic into semiconductors.

What comes next?

The competitive question is execution: whether Tesla can move from a founder's declaration to a funded, staffed, operating fab, in a market where even incumbent foundries spend years and tens of billions of dollars bringing a single gigafab online. Until Tesla names a site, a budget or a timeline, the Terafab remains a stated intent — and the burden of proof sits squarely with the company.