Musk Says Terafab’s Trillion-Dollar Cost Is a Multiyear Plan

A $1 trillion estimate for Elon Musk’s Terafab chip plant has no financing plan behind it, and Musk conceded that the figure will probably materialize over time.

Ron Baron put that estimate on the table at the Baron Investment Conference on November 6, 2026. Baron said Terafab, the chip venture of Tesla and SpaceX, would cost $1 trillion. Musk, appearing by video, agreed that it probably would over time, while noting that spending would occur in phases.

That admission came from transcripts of the conference livestream, not a published financing package, construction schedule or commitment from investors. It leaves the company’s most ambitious claim without a clear source of capital.

The stakes extend beyond Musk’s companies. Terafab would compete with Nvidia and Intel, both of which face a new customer-backed manufacturing effort tied to Tesla and SpaceX.

What has Musk committed to building?

Musk said Terafab aims to produce a terawatt of AI chips annually, a goal he called “terrifyingly large.” He estimated global AI chip output at around 20 to 25 gigawatts in 2026 and roughly 35 to 45 gigawatts in 2027.

Musk said a terawatt would represent about 30 times 2027 production. His own estimated range implies 22 to 29 times, making the comparison approximate rather than precise.

The schedule begins modestly. On SpaceX’s October 20 earnings call, Musk targeted the first SRAM test chip for May or June 2027. SRAM means static random-access memory, a relatively simple type of memory chip used to verify that a manufacturing line works.

Musk set volume production for the third quarter of 2028. He called both dates “very aggressive.” Neither milestone has yet produced a chip.

SpaceX currently buys Nvidia GB300 accelerators at a rate of roughly 200,000 per month. Nvidia, which the source values at $4.51 trillion, reports its results on November 17.

The takeaway: Terafab has a headline capacity target and two tentative production dates, but no demonstrated chip or completed manufacturing line.

How much does the first phase cost?

Texas Gov. Greg Abbott’s office estimated on August 6, 2026 that Terafab’s first phase in Gibbons Creek, Grimes County, would cost more than $16.8 billion. The site would cover 100 million square feet.

An earlier published estimate put Terafab’s total cost at more than $100 billion. That figure precedes Musk’s November 6 acceptance of Baron’s much larger estimate.

The $16.8 billion phase remains the only official cost estimate cited in the source. No disclosed funding plan covers the later stages needed to approach $1 trillion.

Terafab also has a staging plan. It would use a research fab at Gigafactory Texas in Austin to test chip processes, then transfer successful processes to Grimes County.

Musk said this arrangement could shorten a process cycle from about one year to about one month. A process cycle is the time needed to run a manufacturing change through a production line and determine whether it improves yield, the proportion of working chips on each wafer.

Shorter cycles would allow the fab to identify defects, correct them and reduce costs more quickly. Musk nevertheless acknowledged that the plan could fail.

The takeaway: the only official cost figure covers a first phase, while the trillion-dollar headline rests on an unfinanced sequence of later phases.

Who pays for the later phases?

Tesla’s financial position complicates a larger commitment. The company’s third-quarter free cash flow fell 61.63%, while capital expenditure rose 110.5%. Tesla trades at a price-to-earnings ratio near 373 times after operating income fell 55.79% in the quarter.

SpaceX earned $4.40 billion in the third quarter. It nevertheless guided fourth-quarter capital expenditure to $35 billion to $40 billion.

Either company could fund additional Terafab spending through equity issuance, which would dilute existing shareholders. The source does not identify which company will carry that burden.

Special committees at Tesla and SpaceX have discussed combining the companies without agreeing on terms. Their lack of an agreement leaves the ownership and funding structure unresolved.

Terafab therefore adds an open-ended claim on Tesla and SpaceX capital before the companies settle how they might combine. The decision about who pays comes before any decision about how much the project will ultimately cost.

The takeaway: both backers face rising spending demands, and neither a dilution plan nor a final combination agreement answers the financing question.

What does Intel’s participation change?

Intel said in April that it would join Terafab. The arrangement gives Intel’s foundry operation a high-profile customer, but it also means a foundry unit that lost $1.85 billion in the third quarter will help build a rival.

Intel’s 7.75% share gain to $56.30 came after a post-election selloff, and no source linked that move to the Terafab announcement. The partnership nevertheless changes Intel’s strategic position because Terafab is designed around Tesla and SpaceX demand.

A foundry manufactures chips for other companies rather than selling finished processors under its own brand. Intel’s participation would place its manufacturing expertise inside Musk’s vertically integrated effort to control AI chip supply.

The takeaway: Intel gains a prominent customer while helping establish a competing foundry, a trade-off that its quarterly loss complicates.

Did the announcement lift Tesla and SpaceX?

Tesla rose 4.22% to $358.20 after the remarks. SpaceX gained 8.53% to $134.16, while Intel climbed 7.75% to $56.30.

No source tied those gains to Musk’s comments. All three stocks had fallen in a post-election selloff before the conference, so the price moves do not establish investor acceptance of the $1 trillion target.

The source did not provide trading dates for those closing prices. It also did not identify the market capitalization, revenue or profit figures for SpaceX and Terafab, which prevents a fuller assessment of the proposed project’s burden on either business.

The wording of Musk’s answer also matters. He did not announce $1 trillion in committed spending. He accepted the estimate as a likely long-term cost, with expenditures spread across phases that the source does not fully schedule or finance.

The takeaway: the stock moves lack a demonstrated link to the announcement, and they cannot substitute for a disclosed funding plan.

What would validate the trillion-dollar target?

Terafab would need to move beyond its Austin research fab, build the Gibbons Creek site and establish volume production. The first SRAM test chip targeted for May or June 2027 would test the line, while volume production in the third quarter of 2028 would mark a much larger operational step.

Even those dates carry Musk’s own warning. His words “very aggressive” acknowledge the risk of delays, and his acknowledgment that the Austin-to-Grimes County process transfer could fail adds another technical hurdle.

The reported milestones also do not establish that Terafab can manufacture Nvidia GB300-class accelerators at the scale implied by its terawatt target. The source provides no yield data, chip specifications for a completed product, construction budget beyond the first phase or named capital providers for the full plan.

For Nvidia, SpaceX’s current purchasing rate of roughly 200,000 GB300 chips per month shows the scale of the demand Terafab seeks to serve. For Intel, the April partnership tests whether its foundry business can convert that relationship into production.

The takeaway: the project’s credibility now rests on test chips, a working transfer to Grimes County and a disclosed financing structure, not on a trillion-dollar headline.