Terafab: Why Musk Needs His Own Chip Factory
Description
Terafab is Elon Musk's bet that chip supply, not chip design, becomes the bottleneck of the AI decade. Tesla, SpaceX and Intel plan to build a chip fab in Texas: a ~$3B research fab in Austin, then a main campus at Gibbons Creek with filings that add up to about $119B by the mid-2030s. In this video: what Terafab is, how a fab and EUV lithography actually work, the plan and the numbers, TSMC/Samsung/Intel, and an honest bull and bear case. This is analysis, not financial advice.
Chapters
0:00 Hook
0:49 What Terafab is
2:11 How a fab works, and why it's hard
4:19 The plan and the numbers
6:12 The competition
7:38 The bull case
8:33 The bear case
10:01 The thesis
Related
▶ Our full SpaceX breakdown: https://youtu.be/XVSbDtPbv_8
Key sources (every narrated fact has 2+ independent sources in our notes)
• SpaceX Q3 2026 earnings call transcript (Oct 20, 2026): https://stockanalysis.com/stocks/spcx/transcripts/723936-q3-2026/
• SpaceX prospectus (SEC 424B4, risk factors): https://www.sec.
Transcript
Read auto-generated transcript (1910 words)
Kind: captions Language: en Every AI model, every self-driving car, and every [music] humanoid robot runs on the same scarce thing, leading-edge chips. [music] And one company, TSMC, now takes about 72% of the world's contract chip-making revenue. >> [music] >> So, Elon Musk's companies decided to build their own chip factory. It's called Terafab, and its long-term goal is a terawatt of AI compute every year. In March, Musk summed up the logic in one line. We either build the Terafab or we don't have the chips, and we need the chips, so we build the Terafab. This is Bet the Decade, what Terafab is, why chips are so hard to make, the numbers, and an honest bull and bear case. >> [music] >> This is analysis, not financial advice. Terafab was announced on March 21st, 2026, in Austin, Texas. It is a joint project of Tesla and SpaceX, [music] which now owns XAI. In April, Intel joined as a partner. It is meant to make two kinds of chips. One is built for edge inference, the brain inside Tesla's cars and Optimus robots. The other is a high-power chip designed for AI satellites in orbit. The unusual part is the scope. The plan puts chip design, photomasks, logic, memory, advanced packaging, and testing in one place, so engineers can change a design and see the result fast. Why bother? Musk says his suppliers, like TSMC, Samsung, and Micron, are only comfortable expanding at a certain rate, and that rate is far below what his companies want. >> [music] >> SpaceX says the combined chip demand of SpaceX and Tesla is expected to exceed 1 terawatt of compute, which it calls significantly larger than today's global supply. For scale, on SpaceX's October earnings call, Musk said his AI business expects roughly 3 gigawatts of compute online by the end of this year, and about 10 more gigawatts in 2027. A terawatt is a thousand gigawatts. >> [music] >> A chip starts life on a polished disc of silicon called a wafer. Many chips are built on each wafer at once, layer by layer. Each layer goes through a loop. Deposit a thin film, coat it, [music] print a pattern with light, etch it, clean it, and inspect it. Repeat that for dozens of layers, and an advanced wafer can go through more than a thousand process [music] steps. That takes time. A wafer typically spends around 3 months inside [music] the fab, longer on the most advanced processes, before packaging and testing even begin. >> [music] >> The hardest step is printing the finest layers. That takes extreme ultraviolet lithography, or EUV. A powerful laser hits tiny droplets of molten tin, creating light with a wavelength of just 13 and a half nanometers. Because EUV light is absorbed by almost everything, even air, the whole path runs in a vacuum. >> [music] >> Mirrors instead of lenses guide the light onto a pattern mask, and from there down onto the wafer. Only one company in the world makes these machines, ASML in the Netherlands. The most advanced versions cost hundreds of millions of dollars each. Then there's yield, the share of chips on each wafer that actually work. A single tiny defect can kill a chip, and big AI chips have more area to catch one. So, the same wafer can give you mostly good small chips, but far fewer good big ones. New processes start with lower yields, and pushing them up takes months of tuning. Even Intel, with decades of experience, talks about yield improvement quarter by quarter. And the building itself is a machine. A fab runs around the clock inside an ultra-clean room, and it needs huge, steady supplies of electricity and water. That's why the leading edge is such a small club. Tesla already designs its own AI chips, like the ones on this self-driving computer, but outside foundries such as TSMC and Samsung manufacture them. >> [music] >> Step one is a research fab at Tesla's Gig Texas site in Austin. Roughly a $3 billion facility built for a few thousand wafers a month. It will use Intel's 14A manufacturing process. >> [music] >> That makes Terafab a major early customer for 14A, which Intel plans to bring to high-volume manufacturing in 2028. Intel's CEO Lip Bu Tan on Intel's April earnings call, "As we look to continue challenging the status quo, I can think of no better partner than Elon Musk." The main campus goes in Grimes County, east of College Station, Texas, at Gibbons Creek, where a power plant once stood. Process water is planned to come from the Gibbons Creek reservoir. And SpaceX says it will bring its own power, natural gas [music] plants, and very large battery arrays, instead of relying on the grid. The first phase is budgeted at about $16.8 billion. State tax filing sketch four phases running into the mid-2030s, adding up to about $119 billion. >> [music] >> Musk stated output targets are 100,000 wafer starts a month at first, and an ambition of 1 million a month at full scale. And the clock on SpaceX's earnings call on October 20th, Musk said the first test chip, a simple SRAM memory chip, is aimed at around May or June of 2027 at the Austin research fab. >> [music] >> Then he said, "Our aspiration is to reach volume production around Q3 of 2028 at Terafab at Gibbons Creek, >> [music] >> the main Terafab location." He thanked ASML and other suppliers for accelerating equipment deliveries and added, "These are obviously insanely fast time frames for chip production." >> [music] >> So, who is Terafab up against? According to TrendForce, TSMC took a record 72 and 1/2% of global foundry revenue in the second quarter of 2026. Samsung was second with under 6%. TSMC plans to spend 60 to 64 billion dollars on capital expenditure this year alone. >> [music] >> And it has committed at least 165 billion dollars to its fabs in Arizona. Put that next to Terafab's first phase, [music] about 16.8 billion dollars spread over roughly 3 years. Terafab is big for a newcomer, but TSMC spends several times that every single year. Samsung is a Tesla supplier, too. In July 2025, Tesla signed a 16 and 1/2 billion dollar deal for Samsung to make its AI 6 chips at its plant in Taylor, Texas. [music] And Intel is both partner and rival. Its foundry business lost 2.4 billion dollars in the first quarter of 2026 as it races to win outside customers for 14A. Terafab helps Intel make that case. So, Terafab does not replace these companies anytime soon. SpaceX's own prospectus says it expects to keep sourcing a significant of its compute hardware from third-party suppliers. >> [music] >> Here's the bull case to 2030 and beyond. One, demand. If AI compute keeps growing toward Musk's targets, owning chip supply could matter more than any single chip design. Two, speed. With design, masks, chips, and packaging in one place, each design loop could get much shorter. Three, >> [music] >> partners. Intel brings a manufacturing process, and ASML is accelerating deliveries. Four, room to grow. Four phases through the mid-2030s mean Terafab can scale step by step if the early phases work. >> [music] >> In the bull scenario, the Austin fab proves the process in 2027. >> [music] >> Gibbons Creek ramps around the end of the decade, and by 2035 Terafab is a major source of chips for robots, cars, and orbital data centers. That's a scenario, not a forecast. >> [music] >> Now, the honest bear case with data. First, the timeline. [music] Research firm SemiAnalysis estimates phase one needs about 20 months to build, 17 months to equip, and 6 months to ramp. That points to volume production around mid-2030, nearly 2 years after Musk's aspiration. Musk himself put it plainly on the call. "We don't know if we will actually achieve this or not." [music] Second, the deal itself. SpaceX's prospectus says Terafab is in the very early stages, that neither Tesla nor Intel are obligated to remain a part of the project, and that there's no assurance of hitting its objectives within the expected time frames, or at all. Third, experience. Tesla and SpaceX have never run a leading-edge chip fab, and Bloomberg has pointed to Musk's history of over-promising on goals and timelines. Fourth, scale. A terawatt a year is far beyond today's entire AI infrastructure. Local reporting described it as more than 10 times all current AI capacity. Fifth, local friction. Grimes County approved a 100% tax abatement in a four-to-one vote over strong objections from residents. And in September, SpaceX [music] sued to block the release of project records. And the power plant leans on natural gas turbines, which brings its own cost and emissions questions. >> [music] >> So, here is the thesis. Terafab is a bet that in the AI decade, chip supply, not chip design, becomes the real bottleneck and that whoever controls supply controls the pace. The vision is huge and the partners are real. The timeline is aggressive and the risks are written into SpaceX's own filings. [music] Both can be true at the same time. Here's what to watch. The first test chip from Austin around mid-2027. Intel 14A reaching volume in 2028. EUV tools arriving on site. Construction progress at Gibbons Creek and whether Tesla and Intel sign definitive agreements. >> [music] >> Is Terafab the boldest industrial bet of the decade or a step too far? >> [music] >> Tell me in the comments. This is Bet the Decade, not financial advice. >> [music]