Waymo Reveals Custom Compute Behind Its Robotaxis
Waymo’s onboard compute is the brain of the Waymo Driver, turning camera, lidar and radar inputs into driving commands without a human backup. That makes deterministic low latency, durability and redundancy central to the system, not merely benchmark performance. After a decade of co-designing sensors, algorithms and hardware, the Alphabet-owned company’s disclosure offers a rare look at the infrastructure underpinning its commercial robotaxi service and its ability to scale safely.
On Aug. 20, 2026, Waymo unveiled its in-vehicle architecture, saying raw compute power has increased 20-fold in eight years. Its purpose-built 5-nanometer ASIC delivers more than 1,000 TOPS of machine-learning performance for front-end processing, while the latest system handles feeds from 13 high-resolution cameras simultaneously. Two independent compute engines run parallel workloads and provide failover. Waymo named AMD, Micron, NVIDIA, Samsung, Sandisk, Socionext and TSMC among its hardware partners.
All Coverage
1 original reportsThe Backstory
The history behind this eventNo historical echoes for this signal
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.
If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →