Waymo has unveiled a custom 5 nanometre (nm) Application-Specific Integrated Circuit (ASIC) chip.
This strategic move underscores Waymo’s deepening vertical integration efforts, allowing it to fine-tune hardware for its sensor suite and cost structure. The Ojai robotaxi is already deployed in Los Angeles, Phoenix, and San Francisco. Waymo is clearly targeting profitability in the rapidly evolving robotaxi sector with this approach.
Waymo’s vertical integration for autonomous vehicles
Waymo has long been recognised for its rapid expansion in robotaxi services. It consistently broadens its presence in existing cities and enters new markets. This growth strategy hinges on a core requirement: making its autonomous vehicles cheaper to build, operate, and maintain for true profitability.
The company’s sixth-generation self-driving system, which debuted in the Ojai robotaxi, represents a pivotal step. Waymo’s investment in custom silicon illustrates a broader industry trend. Companies are taking greater control over their core technology stack, from software to specialised hardware solutions.
The Ojai robotaxi’s strategic importance
The Ojai robotaxi serves as the primary platform for Waymo’s latest driverless technology. Its deployment across key US cities highlights the company’s confidence. It shows their ability to handle challenging urban landscapes without human backup drivers.
A central tenet of the Ojai’s design is cost-efficiency. Every component, including the newly revealed custom chip, aims to reduce the total cost of ownership and operation. This is critical for scaling a profitable robotaxi service.
Inside the Waymo custom 5nm ASIC chip
The newly developed 5nm ASIC chip is a testament to Waymo’s commitment to custom hardware. This specialised processor is tasked with handling the immense volume of raw data. It processes data generated by the vehicle’s sophisticated sensor array before it reaches the vehicle’s central processing unit.
Consider the Ojai robotaxi, which incorporates 13 high-fidelity cameras alone, alongside LiDAR and radar systems. The chip’s design specifically addresses the need for immediate, high-volume data ingestion and preliminary processing. This is crucial for real-time decision-making.
Performance and data handling capabilities
Waymo states its new custom chip delivers over 1,000 TOPS (trillions of operations per second) of computing performance. This metric positions it competitively with high-end automotive processors, such as Nvidia’s latest DRIVE AGX Thor.
Crucially, this performance is dedicated to machine learning (ML) models and front-end processing. It extracts vital information from raw sensor streams. This includes capabilities like temporal denoising, which significantly improves low-light perception.
Rugged design for real-world operation
Operating a robotaxi fleet means exposing hardware to diverse and often harsh environmental conditions. The Waymo custom chip features a ruggedised design. It is engineered to withstand intense vibration, shock, and extreme temperatures.
Integration with the vehicle’s liquid cooling system ensures optimal thermal management, maintaining performance integrity. Moreover, the chip incorporates built-in redundancy, functioning as two independent engines running in parallel to enhance reliability and safety.
Strategic partnerships fuelling chip innovation
Despite its focus on vertical integration, Waymo hasn’t shied away from critical industry collaborations. The development of this custom silicon involved partnerships with leading technology firms. These include AMD, Micron, Samsung, and TSMC.
Other collaborators such as Sandisk and Socionext also played roles in bringing this complex hardware to fruition. This collaborative approach allows Waymo to tap into specialised expertise. It helps in chip manufacturing and memory solutions, accelerating development and optimising performance.
These alliances highlight the intricate web of development required for advanced autonomous vehicle hardware. Even as companies pursue proprietary solutions, they often rely on a deep ecosystem of suppliers for specialised components and manufacturing capabilities.
Broader implications for autonomous driving
Waymo’s decision to develop its own 5nm ASIC chip reflects a growing understanding within the autonomous vehicle industry. Off-the-shelf solutions may not suffice for the demanding requirements of truly driverless operations. Companies are finding that custom silicon can offer significant advantages in performance, power consumption, and cost.
This trend suggests a maturing of the AV sector. It is moving beyond experimental phases to industrial-scale deployment. Specialised hardware is becoming a differentiator, enabling capabilities that general-purpose processors struggle to match efficiently.
The cost imperative in robotaxi deployment
Profitability has remained a significant hurdle for many autonomous vehicle companies. High development costs, expensive sensor suites, and complex software systems contribute to substantial operational expenditures.
Waymo’s previous generation Driver already achieved a 50% cost reduction. This showcased the tangible benefits of design optimisation. This new custom chip aims to further drive down costs. It makes a scalable robotaxi service more economically viable and brings the prospect of mass adoption closer.
Industry trend towards specialised hardware
Waymo isn’t alone in this pursuit. The push for custom silicon in autonomous systems aligns with a broader industry movement. Tech giants and automotive players alike are investing heavily in proprietary hardware. This allows for greater control over the entire system, from data acquisition to decision-making.
The goal is to eliminate bottlenecks, improve processing speed, and ensure real-time responsiveness. All this happens without the need for a human backup driver. This vertical integration reduces dependence on third-party suppliers for core computing elements, offering a competitive edge and fostering innovation at a foundational level.
Outlook: Waymo’s path to profitability and expansion
The unveiling of Waymo’s custom 5nm ASIC chip positions the company to potentially accelerate its expansion plans. With its sixth-generation Driver now fully autonomous and already deployed, the focus shifts to scaling operations efficiently.
Waymo intends to elaborate further on its compute strategy at the upcoming Hot Chips conference in Palo Alto, California, during the week of August 24, 2026. This event will likely provide more detailed insights into how this custom hardware will underpin their future growth and ambitions.
The company’s ambition extends beyond its current US footprint, with plans for nearly 20 more US cities, alongside potential ventures into international markets such as London and Tokyo. This extensive expansion follows a substantial $16 billion financing round completed in February 2026, which valued the company at $126 billion post-money.
Alphabet remains Waymo’s largest shareholder, demonstrating continued confidence in its autonomous driving division. As the technology matures and costs decline, the industrial impact of Waymo’s approach could resonate globally. It will influence how other companies approach the daunting engineering challenges of driverless mobility, particularly in emerging markets across Africa.
This kind of technological advancement offers potential for significant impact on industrial logistics and public transport infrastructure.
