Waymo is pushing deeper into vertical integration to drive down costs and sharpen its robotaxi performance. The company recently revealed it developed a custom ASIC chip built on a 5 nm process, designed to run preprocessing workloads before data hits the core self-driving system. This enables faster response in dense urban settings—a critical advantage for robotaxis. The disclosure sheds light on what makes Waymo’s sixth-generation system, showcased in its Ojai robotaxi, more cost-effective both in construction and operation. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/))
With the Ojai robotaxi already available to riders in Los Angeles, Phoenix, and San Francisco, Waymo has emphasized that the new platform is cheaper to build, maintain, and run. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/)) The newly disclosed chip contributes heavily to this goal—it handles vast amounts of raw sensory input from the vehicle’s sensors, including 13 high-fidelity cameras, before feeding data through to the vehicle’s central compute system. This lowers latency and power draw. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/))
The chip delivers over 1,000 TOPS of compute performance—trillions of operations per second—placing it in the same performance league as Nvidia’s DRIVE AGX Thor automotive processor. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/)) While the AGX Thor is a known benchmark for high-end autonomous driving compute, Waymo claims its custom silicon achieves unmatched efficiency without sacrificing performance—a crucial trade-off when scaling robotaxi fleets. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/))
Partners and Ecosystem Strategy
Developing the chip in-house doesn’t mean going it alone. Waymo listed several collaborators—including AMD, Micron, Nvidia, Samsung, Sandisk, Socionext, and TSMC—suggesting the project draws on external supply chain expertise even as the core design remains proprietary. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/)) This mix of internal design and external sourcing follows trends in the auto-tech world, where companies optimize performance while leveraging existing chip manufacturing, memory, and sensor infrastructure.
The custom ASIC isn’t the first stroke in Waymo’s efficiency playbook—but it may be its most critical. Prior generations of Waymo’s self-driving hardware have relied on more off-the-shelf components; the new platform signals a shift toward bespoke hardware that can reduce cost per unit at scale. ([techcrunch.com](https://techcrunch.com/2026/08/23/techcrunch-mobility-the-custom-chip-driving-waymos-robotaxi-ambitions/))
As robotaxi deployments expand in more cities, high barriers remain: regulatory approvals, consumer safety, reliability in different geographies, and managing operational costs. Custom silicon can chip away at the last bit of inefficiency—but only as part of a broader stack: sensors, compute, software, mapping, and logistics.
This internal chip development initiative represents a strong signal that Waymo believes full hardware-software integration is essential to turning robotaxi ambitions into commercial reality. What to watch: How Waymo leverages this custom chip in new markets; what trade-offs it makes between performance, safety, and cost; and whether competitors in autonomy follow suit with similar investments in silicon. Scaling isn’t just about where the robotaxis go—it’s about what they’re built on.
