I was on the call when Elon dropped the Terafab timeline and you could feel the room tilt between hope and skepticism. I’ve watched chip fabs rise and stall; this schedule reads like a provocation. If he’s right, the industry will have been hoodwinked into a new pace — if he’s wrong, the bill will be enormous.
Elon says Terafab will begin commissioning machinery in the first half of next year, that a first test chip could appear around May or June 2027, and that the factory would be at scale by the third quarter of 2028. He called that “lightning fast by fab standards.” The project is a joint venture between Tesla and SpaceX — companies that Musk admitted are still figuring out how to split the cost.
Construction crews are already on site in Austin — Why a fab timeline like this demands skepticism
You see cranes and concrete, but a chip fab is more than metal and walls. Bringing a semiconductor plant online means clean rooms, particle control, and equipment supply chains that routinely take years to line up. I’ve covered fabs where a missing valve or delayed lithography tool set schedules back by quarters; this is not a sprint, it’s a choreography.
Building a fab in the United States adds friction: labor costs, regulatory hurdles, and a local supplier base that’s smaller than Taiwan’s. TSMC’s Phoenix project showed how costly and slow U.S. fabs can be in practice — and reporters at The New York Times and Tom’s Hardware detailed those exact pains. Moving this fast is like trying to build a cathedral with a stopwatch.
When will Terafab produce its first chip?
According to Musk, the window for a first test chip is May–June 2027; real production is “late ’28.” A test chip is a proof of concept — it tells you whether the process chemistry and the tools are talking to each other. You should treat that date as the opening of a curiosity loop, not the finish line: test wafers are one thing, sustained high-yield mass production is another.
Equipment arrives by truck and crate — Who is paying for Terafab and how much will it cost?
On the call, Musk said the partners must decide how to split the bills. He warned the initial capital through early 2028 won’t be less than “tens of billions,” and that long-term spending will probably be north of $100 billion (≈€93 billion). Those are headline numbers that shape strategy: $100 billion (≈€93 billion) could buy several advanced fabs, or a fleet of rockets and cars.
If you’re trying to judge risk, ask who guarantees those checks. Musk hinted that the burden will sit between Tesla and SpaceX — two companies with very different revenue profiles and investor expectations. A misread here could leave one partner holding a heavy financial bag.
How much will Terafab cost?
Musk’s public estimate: “not less than tens of billions” through early 2028, and probably “north of $100 billion (≈€93 billion)” over the long term. Translate that into planning: tens of billions could be ≈€19–€37 billion if you assume $20–$40 billion; the $100 billion marker signals multi-decade ambition and massive capital exposure.
A high-volume line is already humming in Fremont — What Terafab means for Optimus and Tesla’s product roadmap
Fremont is shipping early units on a “high volume line,” Lars Moravy said — a tangible step. Musk tied Terafab to Optimus, Tesla’s humanoid robot project, suggesting the factory will be “an incredibly important asset, if not a necessary asset,” to scale those robots.
Electromechanical design remains a limiting factor, especially robot hands. Musk expects that to be solved by the end of 2027, with customer deliveries to industrial buyers following. Pushing a fab online on a 2027–2028 cadence, though, is like threading a needle at Mach 2: doable in theory, terrifying in practice.
Can Tesla and SpaceX share a fab?
Technically, yes — chips for rockets and cars can be made on overlapping process nodes — but the business and governance side is messy. SpaceX and Tesla have different reliability specs, supplier relationships, and cash-flow patterns. Sharing means negotiating priorities when tools break, when equipment queues lengthen, or when specialized masks are needed.
Cybercabs roll through Austin — Regulators are the quiet variable that can upend the plan
Two-seater Cybercabs are operating in Austin and Tesla says customers can buy them this year, though Musk expects most to remain company-owned. That operational reality sits beside growing regulatory scrutiny: the NHTSA is probing Tesla over reports of vehicles speeding through school zones, and European regulators have been vocal about self-driving claims.
Regulatory friction has a cascade effect on strategy. If you want to run robotaxi fleets at scale, you need predictable rules about safety certification and vehicle permits. If regulators draw hard lines, the market Musk describes — “do the majority of passenger miles where allowed” — becomes conditional rather than inevitable.
I’ll watch the commissioning schedule, the arrival of lithography tools, and how Tesla and SpaceX carve up the tabs. You should too — because Terafab is a bet that could reshape supply chains, corporate balance sheets, and the race for autonomy; will Musk’s timetable be a record-breaking sprint or an expensive lesson in optimism?