Nvidia is establishing itself as the foundational technology supplier for the robotaxi industry, with its open platform now powering every major commercial autonomous vehicle program globally. The company's strategy focuses on providing a complete stack for AI training, simulation, and safety validation rather than selling a single chip. This approach arrives as driverless fleets already operate on busy city streets, and as the robotaxi market is projected to reach $400 billion by 2035, with more than 6 million commercial vehicles expected in operation.

How Does Nvidia's Robotaxi Platform Work?

Nvidia's architecture divides into three distinct computing layers, each serving a critical function in autonomous vehicle development and deployment. The training layer runs on Nvidia's DGX systems, which provide the computational power needed to develop and refine autonomous driving models. The simulation and validation layer tests these models in virtual environments before real-world deployment. The third layer sits inside the vehicle itself, handling real-time perception, reasoning, and planning decisions.

For the vehicle itself, Nvidia points to DRIVE Hyperion 10, which it describes as a level-4-ready reference architecture. This design pairs two DRIVE AGX Thor systems-on-chip, built on Nvidia's Blackwell platform, to run vision-language-action models for perception, reasoning, and planning. The system integrates 14 high-definition cameras, nine radars, three lidars, and 12 ultrasonic sensors, all fused together to provide 360-degree coverage of the vehicle's surroundings.

Which Companies Are Already Using Nvidia's Platform?

The adoption of Nvidia's robotaxi stack spans a remarkably broad ecosystem across multiple continents and vehicle types. Uber is scaling a DRIVE Hyperion fleet and plans to reach 28 cities by 2028, collaborating with partners including Lucid, Mercedes-Benz, Nissan, Nuro, Stellantis, Wayve, and Zoox. May Mobility intends to run ride-hailing services on the Uber network using DRIVE, while its vehicles already operate on Lyft's network in Atlanta with the same platform. Bolt is using Nvidia technology to develop and scale autonomous vehicles across Europe, and WeRide plans to bring its GXR vehicle to Southeast Asia through a Grab partnership.

• Waymo: Working with Nvidia on an autonomous computing system for its robotaxi operations
• Wayve, Nissan, and Uber: Developing a global robotaxi program around a prototype vehicle using Nvidia's platform
• Zoox: Relies on DRIVE for in-vehicle computing along with cloud training and simulation capabilities
• Momenta: Builds its software stack on DRIVE AGX running DriveOS for autonomous vehicle control
• Pony.ai: Created a new domain controller using DRIVE Hyperion and DRIVE AGX Thor
• Tensor: Developing a level 4 Robocar with eight DRIVE AGX Thor chips inside the vehicle
• Lenovo: Supplying a DRIVE AGX Thor-based level 4 domain controller for the SWM robotaxi program

This ecosystem demonstrates how thoroughly Nvidia's platform has penetrated the autonomous vehicle industry, from established automakers to specialized robotaxi companies and technology partners.

What Software and Tools Support the Hardware?

The tooling side of Nvidia's platform matters as much as the hardware itself. Nvidia's Alpamayo portfolio bundles open reasoning models, simulation frameworks, and datasets, with the reasoning models specifically aimed at the long-tail problems that trip up autonomous driving in edge cases. Omniverse NuRec rebuilds real-world driving scenarios from sensor data, while Cosmos world foundation models generate physically based variations of those scenarios, both running on RTX PRO Servers. The AlpaSim framework trains and evaluates reasoning-based driving models for improved performance.

Nvidia also offers physical AI datasets, reinforcement learning blueprints, and distillation recipes to help developers optimize their models. Safety is handled through Halos and its Halos OS, which the company says spans inspection, validation, simulation, and continuous testing from cloud to car. Alpamayo 2 Super became available for commercial use on Hugging Face, shipping under the OpenMDW-1.1 license that covers fine-tuning, derivatives, and commercial redistribution. Nvidia says it tops LingoQA in a field of nearly 40 models, and the Alpamayo family has passed 500,000 Hugging Face downloads, indicating strong adoption among developers.

Why Is This Platform Strategy Significant for the Industry?

Nvidia's decision to build an open platform rather than a proprietary system represents a fundamental shift in how the company approaches the autonomous vehicle market. By providing a complete stack that covers training, simulation, validation, and in-vehicle computing, Nvidia has created a standard that multiple competing companies can build upon. This approach reduces fragmentation in the industry and allows companies like Uber, Waymo, and others to focus on their unique algorithms and business models rather than reinventing the underlying infrastructure.

The timing is critical. Driverless fleets are already carrying passengers on busy streets in multiple cities, and the market is moving from experimental pilots to commercial scale. With the robotaxi market projected to reach $400 billion by 2035 and more than 6 million commercial vehicles expected in operation, the infrastructure decisions made now will shape the industry for decades. Nvidia's ecosystem now spans Asia, Europe, the Middle East, and North America, positioning the company as the foundational technology provider across geographies and vehicle types.

Nvidia's chief executive, Jensen Huang, has predicted that 2026 will be physical AI's ChatGPT moment, suggesting the company sees autonomous vehicles and robotics as the next major frontier for artificial intelligence applications. The breadth of adoption across competing robotaxi companies, from Uber to Waymo to regional players like WeRide, suggests that Nvidia's open platform strategy is succeeding in becoming the industry standard.