The News: Samsung is preparing to build a second chip fabrication facility at its Taylor, Texas semiconductor campus — the same site slated to produce Tesla's next-generation AI6 (HW6) chip.
Why It Matters: This expansion signals that demand for Tesla's AI6 chip is serious enough to warrant a second dedicated fab, putting Tesla's autonomous driving and AI hardware roadmap on firmer footing.
Source: @SawyerMerritt on X
Samsung Is Doubling Down on Tesla's AI6 Chip — A Second Texas Factory Is Coming
Samsung isn't just building one chip factory for Tesla. It's planning two. According to reporting from Sawyer Merritt, Samsung is preparing to construct a second fabrication facility at its semiconductor cluster in Taylor, Texas — the same campus where its flagship $25 billion fab is already being built to manufacture Tesla's future AI6 (HW6) chip. For Tesla owners watching the company's autonomous driving ambitions, this is one of the most concrete signs yet that Tesla's next hardware generation is on a serious industrial trajectory.
📊 Key Figures
| Metric | Value | Context |
|---|---|---|
| Total Samsung Taylor Investment | $37 billion | Up from initial $17B commitment |
| U.S. Government CHIPS Act Subsidy | $4.75 billion | Federal support under CHIPS and Science Act |
| Tesla–Samsung Manufacturing Deal Value | $16.5 billion | Multi-year agreement through Dec 31, 2033 |
| AI6 Chip Process Node | 2nm (SF2P) | Samsung's most advanced process technology |
| First Fab Operational Target | End of 2026 | Mass production may slip to early 2027 |
| Max Campus Capacity | Up to 10 fabs | Taylor campus total buildout potential |
What's Actually Being Built in Taylor, Texas
Samsung's Taylor campus is shaping up to be one of the most consequential semiconductor sites in North America. The first fab — a $25 billion facility — is already under construction and is expected to reach full operational status by the end of 2026, though some industry analysts have flagged the possibility of mass production slipping into early 2027. Pilot operations have begun, and critical equipment testing — including extreme ultraviolet (EUV) lithography systems — was scheduled to commence in March 2026.
The Taylor campus is large enough to eventually host up to 10 fabrication facilities, so a second fab was always part of the long-term vision. What's notable now is the timing: Samsung is moving to initiate that second facility while the first is still ramping up. That's a signal of confidence — both in the technology roadmap and in the anchor customer anchoring the whole operation: Tesla.
The Tesla AI6 (HW6) Deal: What We Know
Tesla and Samsung have a formalized, multi-year manufacturing agreement valued at $16.5 billion (approximately 22.8 trillion Korean won), running through December 31, 2033. Under this deal, Samsung will produce Tesla's AI6 chip — also referred to as HW6 — at the Taylor facility using Samsung's 2-nanometer SF2P process node. That's Samsung's most advanced manufacturing process, placing Tesla's next-generation AI hardware at the cutting edge of what's currently achievable in chip fabrication.
To put the 2nm node in context: current Tesla vehicles run on HW4 hardware, which uses a more mature process. The jump to a 2nm chip represents a generational leap in compute density and power efficiency — exactly what Tesla needs to support more capable autonomous driving software and its broader AI ambitions, including the Robotaxi fleet and Optimus robot.
🔭 The BASENOR Take
Timeline: First fab operational end of 2026 (possible slip to early 2027) · Second fab in planning/preparation phase
Impact Level: 🔴 High — directly affects when HW6 vehicles reach owners
Confidence: High — backed by verified deal terms and Samsung's public investment disclosures
Here's the honest read: the second fab announcement matters more as a signal than as an immediate deliverable. Samsung building a second factory at Taylor tells us two things. First, Tesla's AI6 chip demand projections are large enough that a single fab — even a $25 billion one — isn't sufficient. Second, Samsung is treating Tesla as a long-term anchor tenant, not a one-off customer.
For current Tesla owners, the practical question is: when does HW6 land in vehicles? The first fab targeting end-of-2026 for operations means chip production at meaningful volume is realistically a 2027 story. Tesla would then need time to integrate HW6 into vehicle production and begin deliveries. A reasonable expectation is that HW6-equipped vehicles start shipping sometime in 2027, with the second fab providing the production headroom to scale that rollout.
The broader picture here is about Tesla's AI infrastructure becoming genuinely domestic. A $16.5 billion chip deal, manufactured in Texas, using American-subsidized fabs — this is Tesla vertically integrating its AI hardware supply chain in a way that reduces geopolitical risk and gives the company more direct control over its most critical component. That's strategically important in a world where semiconductor supply chains have proven fragile.
For owners currently on HW4, there's no action required. But this news is a meaningful data point that Tesla's next hardware generation has serious industrial backing — and that the autonomous driving capabilities it's designed to unlock are moving from roadmap to reality.
