Samsung's Texas fab enters production for Tesla's AI5 chip on 2nm | Electrek

Skip to main content

Toggle social menu

Toggle dark mode

Search for:

Submit

Toggle search form

Toggle dark mode

Tesla

Samsung’s Texas fab enters production for Tesla’s AI5 chip on 2nm

Fred Lambert | Jul 13 2026 - 6:09 am PT

9 Comments

Samsung has completed its own version of Tesla’s AI5 self-driving chip and is preparing to start production at its foundry in Taylor, Texas, on the company’s latest 2-nanometer process, according to industry officials.

The move confirms Samsung is building AI5 on 2nm — a node that had been expected to debut with Tesla’s next-generation AI6 chip, not this one.

What Samsung revealed

The disclosure came from James Kim, a principal engineer at Samsung Foundry, who wrote on LinkedIn that “the Tesla-Samsung AI5 chip has reached tape-out” and is “scheduled to be manufactured at the Taylor fab using our latest 2nm process and will soon be integrated into Tesla’s newest products.” Kim deleted the post after it was picked up by Korean news reports.

Tape-out marks the point at which a chip’s design is locked and handed over for manufacturing. From there, the design goes through prevalidation, photomask production, and wafer fabrication to create engineering samples, which must pass qualification with the customer before entering mass production.<br>Advertisement - scroll for more content

We reported in April that Musk said Tesla had taped out AI5 to both Samsung and TSMC, with each foundry producing “slightly different versions” of the chip because, in his words, “they translate designs to physical form differently.” Kim’s post indicates Samsung has now fit that design to its own process and is ready to produce actual silicon.

Musk has already shown off a Samsung-made prototype. The AI5 chip he posted on X carried the marking “KR 2613,” indicating it was made in Korea in the 13th week of 2026 — evidence that Samsung’s engineering samples were already coming out of its Korean lines before the Texas fab took over.

The 2nm surprise

The detail that stands out here is the process node.

The working assumption across the industry — and our own prior reporting — was that Samsung’s 2nm line was tied to the follow-on AI6 chip, while AI5 would be built on more mature nodes at TSMC. Kim’s post says AI5 itself is going on 2nm at Taylor.

That matters because Samsung’s 2nm yields have been the gating problem. We reported earlier this year that AI6 had already slipped roughly six months on Samsung 2nm yield issues, pushing its mass production to Q4 2027 at the earliest. Putting AI5 on the same node supports market assumptions that Samsung’s 2nm yield has now crossed 60% — the rough threshold where the process becomes viable for a high-volume customer like Tesla.

Samsung held an equipment installation ceremony at the Taylor fab in April and has said it expects volume production of Tesla’s AI chips there in the second half of 2027.

More at stake for Samsung than Tesla

For Samsung, landing Tesla’s AI5 on 2nm is a proof point it badly needs. The company has trailed TSMC on advanced-node yield for years, and Taylor was a multibillion-dollar bet that has been searching for anchor customers.

There are signs it’s starting to pay off. Recent reports suggest Anthropic could manufacture its own AI chips through Samsung Foundry, and improving yields are exactly what would let Samsung pull in more customers of that caliber. Samsung declined to comment on anything related to its customers.

Electrek’s Take

If Samsung is confident enough to run AI5 on 2nm at Taylor, its yield story is improving, and that’s good for anyone who wants a real second source to TSMC at the leading edge.

For Tesla, though, the same caveat applies as in April: tape-out and early production are milestones, not the finish line. Musk himself has said Tesla needs “several hundred thousand completed AI5 boards line side” before it can switch vehicle production over, and that volume isn’t expected until mid-2027. Engineering samples coming off a Texas line in 2026 don’t change that.

But from Tesla’s point of view, it doesn’t need AI5 to achieve unsupervised autonomy at scale. Of course, we heard that about HW3 before, and now AI4 and the company have yet to achieve unsupervised autonomy at scale.

And the bigger pattern still stands. AI5 is arriving nearly two years after Musk first said it would be in vehicles, the Cybercab is launching on old AI4 hardware, and the AI4.5 stopgap computer quietly added to new Model Ys exists precisely because AI5 kept slipping.

If you’re a Tesla owner, powering your EV with home solar is one of the smartest ways to lock in low fuel costs. With electricity rates climbing nearly 10% last year, home solar protects you against future rate increases. And with lease and PPA options, you can go solar with zero upfront cost and start saving immediately. If you want to find the best deal, check out EnergySage. It’s a...