Waymo, the self-driving car unit owned by Alphabet, has developed its own custom silicon chip to power its robotaxi fleet. The company announced the chip on Thursday, saying the in-house processor will sharpen the vehicles' ability to navigate complex city streets by handling the massive flow of camera, lidar, and radar data generated inside its driverless cars.

News of the announcement circulated widely on X, where the finance account Wall St Engine highlighted the development:

WAYMO BUILT A CUSTOM CHIP FOR ITS ROBOTAXIS $GOOGL 's Waymo has developed its own 5nm ASIC capable of 1,000 TOPS to process the massive flow of camera, lidar and radar data inside its driverless cars. The chip handles front-end sensor processing and AI workloads before feeding… pic.twitter.com/W1U69YaXeL

— Wall St Engine (@wallstengine) August 20, 2026

The new processor is an application-specific integrated circuit, or ASIC, meaning it is built for a specific task rather than general-purpose use. Waymo says the chip can perform more than 1,000 trillion operations per second, putting it on par with Nvidia's latest autonomous driving systems, and that it handles front-end sensor processing and AI workloads before the data is passed along to the rest of the vehicle's systems.

Until now, Waymo has relied on chips from outside suppliers such as Nvidia and AMD. Bringing the design in-house allows Waymo to process sensor data faster and to run the AI models that help its cars read and respond to the world around them.

The chip is manufactured by Taiwan Semiconductor Manufacturing using a 5-nanometer process. While that is not TSMC's most cutting-edge method, it is still widely used by major technology companies.

A New Vehicle for a New Strategy

At the same time, Waymo has opened its next-generation robotaxi, called the Ojai, to all riders in Los Angeles, Phoenix, and San Francisco. The Ojai is now being fitted with the new custom chip as it enters commercial service.

The Ojai is built on a minivan platform made by Zeekr, a brand owned by China's Geely Holding Group. Waymo partnered with Zeekr in 2021 and has spent years developing the vehicle for commercial use. The base vehicles arrive in the United States without Chinese connected-car technology, and are then sent to Waymo's Arizona facility, where the company installs its self-driving system.

The Ojai carries Waymo's sixth-generation autonomous driving system, which is designed to work across multiple vehicle types. The vehicle also includes a redesigned interior and Google's Gemini AI, which acts as an assistant for riders. Waymo currently has around 300 Ojai vehicles in its commercial fleet, and riders in the three launch cities may be matched with one when they book a ride.

Costs and Expansion

Waymo plans to expand the Ojai to Denver, Las Vegas, and San Diego later this year. The company operates robotaxis in 11 US cities in total and has said it provides more than 250,000 paid rides each week, which makes the per-vehicle economics of its hardware a meaningful factor as the fleet grows.

The Ojai is designed to be cheaper to build, run, and maintain than the older Jaguar I-Pace vehicles Waymo has used for years. The goal is to help Waymo move toward profitability at larger scale. However, US tariffs on Chinese-built vehicles have raised the cost of each Ojai imported into the country, since the vehicles are manufactured in China before being fitted out in Arizona. That adds pressure to Waymo's cost targets.

Waymo remains the leading robotaxi operator in the United States, ahead of rivals including Tesla and Amazon's Zoox. In-house silicon is already an established practice among autonomy developers: Tesla, which relies on camera-based systems rather than lidar, has designed its own driver-assistance computer chips since 2019. Waymo's move extends that approach to a fully driverless, fare-carrying fleet, and the combination of a purpose-built vehicle and a custom chip represents a shift toward a more vertically integrated model for the company.

Waymo's silicon effort also fits into a broader Alphabet strategy. Google has been building its own Tensor Processing Units for AI workloads since the mid-2010s, and Amazon and Microsoft have pursued custom data-center chips as well. Waymo's chip program forms part of that wider effort to reduce dependence on outside suppliers such as Nvidia, whose AI chips remain in demand across the industry. As the Ojai rollout extends to new cities, the growth of the chip-equipped fleet will offer the clearest measure of how the strategy affects Waymo's costs.