TLDR
- Elon Musk announced the Terafab project launches in 7 days, a chip manufacturing facility at Tesla’s Austin Gigafactory
- The factory is designed to address a potential chip shortage Musk sees limiting Tesla’s growth in 3–4 years
- Tesla’s long-term Optimus robot target could require 200 million+ chips per year — 50x current demand
- Morgan Stanley analyst Andrew Percoco estimates the fab could cost $35B–$40B and produce chips by 2028
- Musk cited geopolitical risks as a key reason for building chip capacity domestically
Tesla stock rose 0.6% in early Wednesday trading to $401.75, after CEO Elon Musk confirmed the company’s “Terafab” chip factory will launch on March 21 at its Austin, Texas Gigafactory.
Musk first hinted at the project in a January 2026 investor call, and confirmed the Austin location in a March 14 post on X. The facility will sit within Tesla’s sprawling 2,500-acre Gigafactory Texas campus.
The AI industry is about to run out of chips. The world's largest foundries are maxed out and the demand curve is still going vertical.
Tesla’s TeraFab is a bet on a solution no other company in the world has been ambitious enough to try…. build the chips yourself
Quick math… https://t.co/EXMHu5GYbC pic.twitter.com/31Ln43JLiO
— Josh Kale (@JoshKale) March 14, 2026
The core reason is simple: Musk thinks Tesla is going to run out of chips. “When I look ahead and say what’s the limiting factor for Tesla growth, if you go 3 or 4 years out, I think it actually is chip production,” he told investors in January.
He also flagged memory and AI logic capacity as potential constraints. “Is there enough AI logic and enough memory, enough RAM for our volume?” Musk said on the call.
Why Tesla Is Building Its Own Chips
Morgan Stanley analyst Andrew Percoco laid out the numbers. If Tesla hits its long-term Optimus humanoid robot target of 100 million-plus units per year, it would need over 200 million chips annually. That’s more than 50 times Tesla’s current chip demand across its auto and robotaxi lines.
Percoco said Tesla’s decision to build internal chip capacity is tied to two things: geopolitical concerns and the Optimus program. Management flagged that AI compute could become a bottleneck within three to four years.
Musk was direct about the geopolitics angle. He said the Terafab needs to include “logic, memory and packaging domestically” to protect against supply chain risks. “I think people may be underweighting some of the geopolitical risks,” he said.
Building a fab is not cheap or quick. For reference, Micron’s Boise memory chip facility broke ground in 2022 and isn’t expected to produce chips until 2027.
Cost and Scale of the Project
Percoco estimates Tesla could face a bill of $35 billion to $40 billion for its own chip fab capacity. Even in an optimistic scenario, he says chips wouldn’t come out until 2028.
That’s a stretch from Tesla’s typical capital spending. The company normally spends under $10 billion a year on plants and equipment, though it has said it plans to spend $20 billion in 2026 as it scales up its robotics ambitions.
Percoco called the fab a “Herculean task” and noted Tesla could also look to partner with existing chipmakers rather than go it alone. He rates Tesla stock Hold with a $415 price target.
Tesla stock came into Wednesday’s session down 11% year-to-date but up 77% over the past 12 months. The S&P 500 and Dow Jones futures were both up slightly on the day.
The Terafab is set to officially launch on March 21.





