Breaking News: Tesla's Revolutionary Upgrades Unveiled!
January 23, 2026
Get ready for some exciting revelations from the world of Tesla! Today, we dive into the latest developments that are set to revolutionize the automotive industry.
Tesla's Three-Chip Revolution: Unlocking New Possibilities
Recent discoveries in Tesla's Electronic Parts Catalog have sparked curiosity and intrigue. It seems that Tesla is gearing up to release, or has already begun utilizing, a new HW 4.5 FSD computer in its vehicles. While the name suggests a minor revision, the technical advancements could be a game-changer.
The Eagle-Eyed Discovery
Todd deRego, an observant Tesla enthusiast, spotted a new component listed in the official catalog. This part, named "CAR COMPUTER - LEFT HAND DRIVE - PROVISIONED - HARDWARE 4.5," carries a price tag similar to other FSD computers, indicating a potential replacement or revision rather than a brand-new addition.
The 3-SoC Theory: A Game-Changing Architecture
Longtime Tesla firmware hacker Green has provided an intriguing insight. According to Green, Tesla's firmware has long referenced a 3-SoC (System-on-Chip) design. Historically, Tesla's FSD computers (HW3 and HW4) utilized a dual-SoC setup, offering standard redundancy with two identical chips processing the world simultaneously. However, the potential shift to a 3-SoC architecture opens up a world of possibilities.
Why Three Chips?
The benefits of a 3-SoC setup are twofold. Firstly, it offers increased raw throughput, enabling larger and more complex inference models. As FSD v14 and future models grow exponentially in size, a 3-SoC setup allows Tesla to distribute the inference load across multiple chips, running larger and more intelligent neural networks.
Secondly, it enhances the vehicle's decision-making process. In a 2-chip system, disagreements between computers could lead to disengagement and safety warnings. With a 3-chip system, Tesla can implement Triple Modular Redundancy (TMR), allowing the system to "vote" and continue driving smoothly even if one chip hallucinates an obstacle.
Additionally, a 3-SoC architecture could enable the car to run two chips with standard redundancy while the third chip runs an experimental version of FSD in the background - Shadow Mode. This innovative approach allows Tesla to validate next-gen software in real-world consumer cars with zero safety risks.
A Bridge to AI5: Filling the Gap
With Tesla's next-generation FSD chip, AI5, slated for production later in 2026, there's a potential gap in computing power. Software progress often outpaces hardware, and Hardware 4.5 could be the perfect bridge. By introducing a slightly more capable iteration, Tesla ensures that millions of cars produced before AI5's deployment aren't left behind as neural networks continue to evolve.
Unsupervised Robotaxi Rides: A Historic Milestone
In a momentous development, Tesla has begun offering unsupervised Robotaxi rides to the general public in Austin, Texas. As of January 22, 2026, riders in Austin can hail a Model Y Robotaxi with no one in the vehicle. This marks a significant step towards truly driverless public transit.
Until now, Tesla's Robotaxi tests in Austin included human safety monitors, either in the passenger or driver's seat. However, Tesla has now removed this safety net, allowing riders to experience fully unsupervised rides. Tesla is still visually monitoring these vehicles, with multiple Tesla cars trailing them from a distance.
This launch puts Tesla in an exclusive club alongside Waymo and Zoox, companies operating driverless public transit. Tesla's approach, relying solely on cameras and AI, differs greatly from Waymo's use of expensive LiDAR arrays and pre-mapped HD geofences. If this pilot in Austin continues successfully, it will validate Tesla's strategy and pave the way for further expansion in Texas and beyond.
AI5: Production Insights and Performance Comparison
According to reports from Korea and updates from Elon Musk, Tesla's next-generation chip, AI5, is set to enter production in the second half of 2026. Samsung's foundry in Taylor, Texas, will begin key equipment tests in March to prepare for mass production.
AI5 will be crucial for advancing FSD, as Elon Musk highlighted. Solving AI5 was seen as an existential challenge for Tesla, with Elon personally dedicating his Saturdays to working on it. AI5 is expected to deliver impressive capabilities, with a single SoC equivalent to NVIDIA's H100 (Hopper) and a dual AI5 system equivalent to NVIDIA's B100/B200 Blackwell.
This performance comparison puts AI5 on par with NVIDIA's $30,000 server chip, designed for chilled data centers. Tesla's achievement is remarkable, as AI5 will be a compact, low-voltage car battery-powered supercomputer equivalent, fitting behind the glovebox.
With AI5 architecture finalized, Tesla's silicon design team is already looking ahead to Dojo 3 (AI6). While AI5 will power the vehicles, Dojo will handle the server and inference training. The focus is now shifting to the hardware that teaches those neural networks.
Stay tuned for more exciting updates from the world of Tesla!
