Samsung Foundry has formally clinched a gargantuan $16.5 billion chip agreement with Tesla, one of its biggest deals to date. A Tesla founder and CEO Elon Musk took to X (the artist formerly known as Twitter) to say that the Texas fab of Samsung will build the Tesla’s next-gen chip, the A16 AI.
The Samsung Foundry division was recently in the news”:/> for its advanced nodes as well as its older ones. Just as production issues plague new tech like the SF2 node, there’s still demand for the old processes at Samsung Foundry, particularly for consumer tech like the Switch 2, which makes use of the 8N node.
And now Samsung Foundry is at the center of Tesla’s next-gen AI plans.
Elon Musk Says Tesla’s A16 Chips Will Be Made by Samsung
Samsung Electronics, Samsung Foundry’s parent company had secured a mega $16.5 billion contract with a world-spanning company, as reported by Reuters. Elon Musk himself disclosed that the company in question was none other than Tesla, and the product in question was the A16 AI chip, for future Tesla models.
New engineering!“” SpaceX’s giddy billionaire boss, Elon Musk, teased last Friday in an allcaps tweet, offering no other details (except for a follow-up tweet that day announcing, “There will be lots more new info from Starbase/NewAcquisition soon“, then nothing else).
The contract, which runs through 2033, also leaves Samsung plenty of time to scale and refine its manufacturing techniques.
Still No Clear A16 Chip Node — Going Older or Leading-Edge?
The process node in question hasn’t been confirmed, but Reuters speculates that it may not be an advanced chip. This has many wondering if Samsung Foundry will be adopting an advanced 5nm or 4nm process node for the A16 chip.
However, contrary reports fromX analyst Jukanlosreve and Chosun Ilbo claims Samsung’s 2nm SF2A node that’s focused on auto-grade chips may be the pick. And due to the terms being so long, Samsung Foundry might start transitioning to its own 2nm SF2A node as the yields of the node improve.
Although it is not public yet in the point of yield, the SF2A node is considered a good result. Samsung’s base SF2 node is said to have 40% yield — commonly used for chips such as the Exynos 2600.
Tesla’s A16 chip will be a bigger die and that fact that timeframes have been extended until 2033 gives Samsung enough breathing room to work out early bugs and improve yield.
Samsung Foundry’s Strategic Win
The landmark Tesla contract puts them head to head with TSMC who was providing the chips for Tesla until now. It also cements Samsung’s foothold in the automotive AI market, a future growth engine for semiconductors.
It may have been Samsung’s track record with consumer devices along with its willingness to tweak older nodes for new jobs that sealed the deal with Tesla.
There’s a New Sheriff in Town — Elon Musk Will Be Running It
As is his wont, Elon Musk pledges to be hands on. He stated he would:
Walk the line yourself to speed up action.
This indicates there is a strong level of integration between the design arms at Tesla and the fabs used to produce Samsung’s A16 chips, all of it to ensure the A16 is capable of delivering the sort of performance and AI-capable capabilities that Tesla requires.
FAQ
How much is the Tesla-Samsung Foundry deal worth?
What chip is Samsung Foundry fabricating for Tesla?
Could Samsung Fab Tesla’s A16 chip on its nextprocess node?
When will Samsung begin making Tesla’s A16 chip?
Why would Tesla pick Samsung over TSMC anyway?
Summary
The Samsung Foundry / Tesla A16 chip agreement is a major change in the global manufacture of chips and underscores Samsung’s growing prominence in AI-driven automotive innovation. With $16.5B committed over several years, both companies are readying the next chapter in the saga of autonomous vehicle technology — and Samsung Foundry is front and center.
