Tesla DOJO 2, DOJO 3 and AI5 chips are critical for Tesla to become an AI hardware company. XAI and Tesla will buy create and build millions of these chips into their data centers and the Tesla products.
XAI and Tesla are buying 1 million Nvidia B200/B300 chips over the 14 months from February, 2025 to April 2026. This is about 20% of Nvidia’s production. If Tesla can make the Dojo 2 chip at scale and volume production in the second half of 2025 this is huge. Tesla has made and installed at least 11,000 Dojo 1 chips but they only have 362 Teraflops of compute. The Dojo 2 chips have ten times the compute.
If Tesla has 25 Dojo 2 chips per wafer and orders 20,000 wafers from TSMC then this would be 500,000 Dojo 2 chips. This would surpass AMD in AI GPU chip production at 400,000 to 500,000 and put Tesla in second place behind Nvidia. Nvidia will make about 5 million B200-B300 chips in 2025.
If Tesla goes to Dojo 3 chips in late 2026 and buys from itself and XAI buys from Tesla this could be multi-million chip order for 2026-2027.
TSMC has described the Tesla Dojo chip process enabling over 40 times the compute performance by 2027. This would put in at about 15 Petaflops or near the performance of an Nvidia B200 at 20 petaflops.
IN April 2024, at TSMC’s North American Technology Symposium, the company says Dojo’s next-generation training tile — the D2, which puts the entire Dojo tile onto a single silicon wafer, rather than connecting 25 chips to make one tile — was already in production.
In July , 2024 AI5 is ~18 months away from high-volume production. This would be end of 2025.
In March 2025, Elon said that the Tesla Dojo 2 chip will be in volume production in a few months. The Tesla Dojo 2 chip has ten times the performance of the Dojo 1. If Tesla Dojo 2 is ten times better than the Dojo 1 training chip then it would be 3.6 petaflops for fp8. This would be several time better than the Nvidia H100 and maybe 40% of the B200.
Tesla Dojo 3 in late 2026.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
The Tesla brand has been severely, and in my opinion irreparably, damaged, and no tech could save it.
This is in part due to Musk’s action, but it is only in part due to that. It is also due to the products: with the speed pedal issue, the not-so-stainless steel exterior, and the not-so-glued panels Cybertruck has been a PR disaster, then there is the supposedly revolutionary Tesla Semi that should have changed the freight business, but it was a blip on the radars and tesla removed the order button from the website. Finally there are the general stats about the fact that teslas have the highest fatal accident rates.
Furthermore, with the 50k downpayment in 2017 for the Tesla Roadster (that in 8 years failed to materialize) and all the issues and recalls in the 100k+ cybertruck, Tesla significantly alienated its more affluent customer base, regardless of the impact of Musk’s image.
Tesla has been instrumental in promoting electric vehicles, but the company is tarnished. In my opinion, they could invent and put Blade Runner-style flying cars on the market tomorrow, and most people would still not buy them, waiting for another, more reliable car company to catch up with the technology.
Welp, TSLA already surpassed the dreams of John DeLorean’s DMC. I’ve never looked at any TSLA product with envy, and I’ve been waiting for the shift from companies being valued for their “Apple-type vibe” to something more traditionally value-based (producing RELIABLE, reasonably priced cars). Perhaps the great uncoupling recession will bring that back into style. Still, established companies (even zombie companies like GM) are struggling to sell products while designing to Congressional/Bureaucratic mandates – they’re not allowed to design for value and the steady trajectory of emissions and collision mandates (for instance), means the tech is never mature before it becomes obsolete. One example is the “peak reliability” of the 1997-2007 GM light truck motors (e.g. the 300,000 mile all day long LS1), which was degraded first by “displacement on demand and VVT” and degraded again in the subsequent LT generation by the complex direct injection fueling system (huge recall in the news recently). There are a lot of NPC bros driving around in turbo Fords and Jeeps, and while the torque curves are impressive when new, on the dyno, a 30psi boosted, 300hp, 2L 4 cylinder with 9-speed transmission is a disposable component. People look at the VW and Cummings diesel scandals (several years ago), easily adopting the conclusion they were provided by the media, that these corporations are unscrupulous, while not entertaining the idea that the regulations are so impossibly strict that these companies basically HAD TO CHEAT to bring a durable, powerful, USEFUL product to market. Pretty sure they’ll be required to inlay airbags in the seatbelts by 2030 /s.
Brian, Dojo 2 is not a single wafer, it’s a system on wafer. Not the same thing. Known good dies on a substrate wafer.
Tesla needs to execute. Solar Roof and CyberTruck were flops. Model Q postponed. Semi allegedly 2x or more $$$
Chinese EV companies are chomping at them.
Their only true advantage is potentially their Robotaxi network. But even if all existing Teslas capable of being Robotaxi’s joined the network, they’d be only a fraction of the size of Uber.
Brian,
Interested in Your take on the recent earning report and where Tesla X is heading.
The earnings numbers came in bad as expected. Q1 was ramping the refreshed model Y. This caused the drop in production and deliveries.
The call re-affirmed that there will be model 3 and Y similar form new lower priced cars. There was also re-confirmation of the June austin unsupervised paid service ride delivery (robotaxi) starting on day one with 10-20 model Y vehicles. This will then expand so by end of 2026 there will be millions of vehicles. This means awakening the fleet in the US for all who want to participate with their Tesla cars. The stock went up about 20+% since the call so the forward looking news was positive. Perhaps this month for lower priced car model and June 1 or days later for Austin robotaxi. Hitting those targets will be huge. I am expecting those things to happen and for new releases of FSD with substantial improvement (3-5X on miles) each quarter this year. There should be licensing of FSD to at least one or two companies. Those major developments should send the stock to over $1000 by the end of 2025. If the robotaxi ramp reaches 100,000+ cars by the end of 2026, the shares could hit $1500. Also helping earnings will be a doubling of energy per quarter to 20+ GWH per quarter.
There is also my expectation that macro will settle down in 2-3 months with some general resolution of the trade situations and that the DOGE cutting and congressional bills pass and the pace settles into something predictable. Expecting $200-350 billion per year in cuts. DOGE claim now is $160 billion. The Ukraine war ending would be another 5%+ boost – lowers energy prices and food prices. Just had the Ukraine minerals deal. If this does not happen or how macro shakes out is a plus or minus up to 20%.