Tesla says millions need Full Self-Driving hardware upgrades after promising they didn't

Tesla says millions will need Full Self-Driving upgrades -- after promising they didn't

Tesla has held an earnings call to discuss its first quarter 2026 results, and revealed more about its plans, including a change of heart on promises that millions of cars would only need software updates for zero-supervision Full-Self Driving. In the Q1 2026 call, CEO Elon Musk said that cars based on Hardware 3 (aka HW3) would need to be retrofitted with both new computers and new cameras. The current generation, Hardware 4 (HW4), arrived for the Tesla Model S, X, and Y in 2023, and the refreshed Model 3 in 2024. HW3 arrived in 2019, and executives have been offering hope for software-only FSD upgrades as recently as October 2025 despite Musk expressing doubts in January that year. The company already started shipping a mildly upgraded HW4 revision with the Model Y in January 2026, and during the call previewed a revision (possibly an AI4.5 Plus or HW4 Plus) that would double RAM to 32GB and deliver an estimated 10 percent improvement to computing power. The upgrade process could be elaborate. Musk said in the call that Tesla was exploring "micro-factories" in major urban areas to avoid overwhelming service centers, where retrofits would reportedly be "extremely slow." He didn't provide a timeline for when these factories might be ready, or when upgrades would begin. The impact could be substantial. Tesla noted in its earnings that there are 1.28 million active FSD purchases and subscriptions, and some customers paid as much as $15,000 for the option. While they won't necessarily have to pay extra, they now face waits for hardware installations even when unsupervised self-driving is ready. Where Tesla is going next in 2026 It's spending billions more this year to pivot to AI Musk also used the Q1 2026 call to outline Tesla's near-term plans, particularly its expansions into AI and autonomy. He confirmed that Cybercab production had begun at Gigafactory Texas, and that the robotaxi was built to meet federal safety standards and won't face the 2,500-car yearly standards exemption cap that limits rivals like Waymo. It can make as many as it likes. The executive also revealed that manufacturing for the third-generation Optimus humanoid robot at its Fremont plant is now expected to start in late July or August. This will come just a few months after the end of Model S and X production, and Musk claimed it would be possible despite the "insanely fast speed" needed to dismantle an existing assembly line and install a complete replacement. The strategy shift will include dramatically larger spending. Musk warned that Tesla's capital expenses (capex) would surge to $25 billion in 2026 to cover its AI efforts, including manufacturing, software, and data centers. That's $5 billion higher than Tesla's previous peak, but Musk argued it was "well justified" for a future where the company relies on more than EVs and energy for revenue. Related US EV sales dropped in early 2026 for nearly everyone except Tesla The end to federal EV incentives was the main factor. Posts 2 By Jon Fingas Tesla will be dependent on EVs in the near future Robotaxi service and robots aren't ready for prime time Musk was tempered about expectations for AI initiatives while discussing Q1 2026 earnings. He expected Cybercab production to be "very slow" at first, with "exponential" growth late in 2026. Likewise, he's no longer predicting Optimus robot numbers where he previously claimed Tesla would make 10,000 units by the end of 2025. The robotaxi service itself is also very small. While existing service in Austin has expanded, Tesla only recently started offering unsupervised rides in Dallas and Houston. San Francisco Bay Area trips still require safety drivers, and the company only said in its shareholder deck that "preparations [are] underway" for expansions to Arizona, Florida, and Nevada. Add in high crash rates and the offering is still well behind Waymo's multi-city footprint. A budget EV may be key As such, Tesla is still heavily reliant on sales of human-driven EVs for its near-term revenue. This includes the long-wheelbase Model Y L that just started testing on U.S. roads. And while Musk didn't acknowledge it during the call, there are rumors that Tesla is back to designing an affordable EV that could reach a wider audience. Subscribe to the newsletter for Tesla pivot analysis Get the newsletter for concise context on Tesla's hardware retrofit decisions and AI pivot, delivering clear, focused coverage that turns complex moves into understandable implications for owners and industry watchers. Get Updates By subscribing, you agree to receive newsletter and marketing emails, and accept our Terms of Use and Privacy Policy. You can unsubscribe anytime. The extra spending on robotaxis and robots is meant to fulfill long-view aspirations, not to balance its books in the next few months or years. Source: Tesla (YouTube)

Tesla announces HW4 Plus with doubled memory -- will HW4 follow HW3 to the grave?

During Tesla's Q1 2026 earnings call, CEO Elon Musk revealed that Tesla is planning an "AI4.1 or AI4 Plus" upgrade to its self-driving computer that doubles the RAM from 16 gigabytes to 32 gigabytes per chip -- taking the total system memory to 64 gigabytes. The announcement came on the same call where Musk confirmed (or rather reconfirmed after backing away from the admision last year) that HW3 "simply does not have the capability" for unsupervised FSD. For the millions of Tesla owners currently driving HW4 cars, the timing raises an uncomfortable question. Musk described the upgrade while explaining why Tesla's next-generation AI5 chip won't go into vehicles anytime soon. Since Tesla believes AI4 is sufficient for unsupervised Full Self-Driving, AI5 will instead go into Optimus robots and data centers. But in the same breath, Musk acknowledged that AI4 hardware will eventually feel its age: "At some point the AI4 hardware is going to get like so old that it's like, okay, the only reason to keep the factory open is for AI4," Musk said. "We are planning an AI4 upgrade to use newer generation RAM. So it'll go from 16 gigabytes to I think 32 gigabytes per SoC. So 64 gigabytes total, and probably a 10% increase in compute and in memory bandwidth." Musk called it "AI4.1 or AI4 Plus" and said it would go into production "next year," dependent on Samsung completing the modifications. Samsung fabricates Tesla's current AI4 chips on its 7nm process. This isn't even the first AI4 revision. In January 2026, we reported that Tesla quietly started shipping a new "AI4.5" computer in the 2026 Model Y built at Fremont. That version appears to use a three-chip design instead of AI4's original two-chip architecture, though Tesla never formally announced it. So the upgrade path is already AI4 → AI4.5 (shipped) → AI4 Plus (announced for 2027). That's three hardware revisions in roughly two years -- for a computer Tesla says is already good enough for unsupervised driving. Here's the pattern that should concern every Tesla owner: Tesla sold millions of vehicles with HW3 from 2019 onwards, telling buyers that every car had "all the hardware needed for full self-driving." The company collected up to $15,000 from customers who purchased the FSD software package based on that promise. On this very same earnings call, Musk finally admitted what we've been reporting for over a year: HW3 can't do unsupervised FSD. It has only one-eighth the memory bandwidth of HW4. Tesla is now proposing to build "micro-factories" to retrofit roughly 4 million cars -- an enormously expensive admission of a broken promise. Now Tesla is making the same claim about HW4: it's sufficient for unsupervised FSD. But the company is simultaneously shipping AI4.5 and planning AI4 Plus with doubled memory and more compute. Memory bandwidth was the specific technical limitation Musk cited for killing HW3. The older chip had only one-eighth the bandwidth of HW4 -- and as Musk put it, "memory bandwidth is the choke point" for AI inference. AI4 currently offers approximately 384 GB/s of bandwidth using GDDR6 memory -- a deliberate upgrade from HW3's LPDDR4 that actually exceeds what NVIDIA's Orin (~205 GB/s) and even the newer Thor (~273 GB/s) deliver. Where AI4 falls short is total memory capacity: 32GB versus Orin's 64GB. The AI4 Plus upgrade specifically addresses that gap, doubling total RAM from 32GB to 64GB and boosting bandwidth further. The fact that Tesla's own engineers see the current memory headroom as tight enough to warrant an upgrade suggests AI4 may not be as future-proof as Musk claims -- even if it handles today's FSD software. The question is whether AI4 can run the software Tesla will need two, three, or five years from now as the neural networks grow larger and more complex to deliver unsupervised self-driving at scale -- exactly the trap that killed HW3. We've watched this movie before. Tesla told HW3 owners their cars had all the hardware needed for full self-driving. We were skeptical for years, and we were right -- Musk just confirmed it on this call. Now Tesla owes roughly 4 million owners either a free retrofit or compensation, and the company's solution is to build entire factories just to handle the upgrades. If you believe that's going to happen, I have a bridge to sell you. Today, Tesla is telling AI4 owners the same thing: your hardware is sufficient. But the company's own actions tell a different story. Musk has been pushing hard the narrative that unsupervised doesn't require AI5, and that's clearly aimed at boosting confidence in HW4 achieving unsupervised driving. He doesn't want Tesla buyers to wait for it, but then why AI4 Plus? And we're not saying HW4 is definitely going to follow HW3 to the grave. AI4 is genuinely more capable hardware, and the gap between AI4 and what's needed for unsupervised FSD may be smaller than the chasm that existed with HW3. But Tesla has burned through all of its credibility on hardware sufficiency claims. When Musk says "AI4 is enough," the only honest response from anyone who's been paying attention is: we'll see.

Tesla Drops Hardware 3 From Unsupervised FSD Path, Offers Upgrade Route

Tesla has officially ruled out Hardware 3 as a platform for Full Self-Driving in unsupervised mode, marking a notable change in the company's long-term autonomy roadmap. During its latest quarterly earnings presentation, CEO Elon Musk said the older in-vehicle computer no longer meets the technical requirements Tesla now considers necessary for unsupervised operation. According to Musk, Tesla once believed Hardware 3 would be sufficient, but further development showed that the gap to Hardware 4 is too large, particularly in memory bandwidth. That bandwidth deficit appears to be central to the decision. Musk said Hardware 3 has only a fraction of the memory bandwidth available on Hardware 4, and Tesla now views that as one of the key constraints for future unsupervised driving capability. In practice, this means existing vehicles equipped with HW3 will not move beyond supervised Full Self-Driving, even if they continue to receive feature updates. For affected owners, Tesla says it will offer two possible paths. One is a trade-in incentive for customers moving to a newer vehicle based on Hardware 4. The second is a retrofit option for existing cars, although the company indicated that this would involve more than replacing the onboard computer alone. Tesla also expects the camera hardware to be swapped, suggesting that both the sensor package and processing platform need to be updated together. Musk added that Tesla may have to create dedicated retrofit facilities in major urban regions to handle these conversions efficiently. He described ordinary service-center execution as too slow and operationally inefficient for this kind of large-scale upgrade work. That implies Tesla is still working out how to industrialize the retrofit process before it can be deployed broadly. The difference between HW3 and HW4 is already reflected in Europe. Tesla recently introduced FSD Supervised in the Netherlands, but for now only Hardware 4 vehicles have gone through testing and approval with Dutch regulator RDW. That leaves HW3 cars outside the currently approved rollout. Tesla's Head of Autopilot and AI Software, Ashok Elluswamy, said a lighter build of FSD V14 for Hardware 3 is expected at the end of June. Tesla says it is intended to retain the main functional set of V14, but it still will not place HW3 on the unsupervised path. As with any deployment in the Netherlands, approval from RDW remains a requirement. The announcement creates a clearer split in Tesla's installed base. Hardware 4 is now the company's active foundation for unsupervised Full Self-Driving ambitions, while Hardware 3 remains limited to supervised implementations. For current owners, that distinction has direct consequences for future software capability, vehicle positioning, and potential upgrade cost or inconvenience. Tesla released the hardware update alongside its latest financial results. The company reported quarterly revenue of $22.39 billion, up 16 percent year over year. Operating profit rose 30 percent to $3.67 billion, while net profit reached $491 million. Tesla delivered around 358,000 vehicles during the quarter. It also said active FSD subscriptions climbed to 1.28 million, representing 51 percent growth compared with the same period last year.