Nvidia is embarking on a transformative partnership with ride-hailing giant Uber and several prominent automakers, aiming to deploy an impressive fleet of 100,000 self-driving taxis and delivery vehicles in the coming years. This ambitious initiative, valued at a potential $4.8 trillion in the computing sector, will see Nvidia supplying its advanced Drive AGX Hyperion 10 platform to automakers. These manufacturers, in turn, will equip Uber with Level 4 autonomous vehicles, capable of operating independently within geofenced areas without human intervention. This significant announcement was made by Nvidia founder and CEO Jensen Huang during his keynote address at the company’s 2025 Global Technology Conference (GTC) in Washington, D.C.

A substantial, though unspecified, portion of these 100,000 vehicles will be designed and manufactured by Stellantis. Production for Stellantis is slated to commence in 2028, with an initial deployment of 5,000 vehicles targeted for the U.S. market. These vehicles will be constructed upon Stellantis’s versatile K0 medium van and STLA Small platforms. To facilitate this complex integration, Stellantis is also collaborating with electronics manufacturer Foxconn, leveraging their expertise in hardware development and vehicle integration.

Beyond the Uber partnership, Lucid Motors is independently integrating the Hyperion architecture into its future vehicle lineup, extending the promise of driverless operation to privately owned cars. This includes its eagerly anticipated $50,000 SUV, likely to be christened the "Earth." Production for this next-generation mid-size SUV is expected to begin at the end of 2026. Initially, it will launch with an advanced hands-off, eyes-off-the-road driver assistance system, which is planned for a subsequent upgrade to a full Level 4 autonomous system, eliminating the need for a human driver. This development complements Lucid’s earlier announcement from July, detailing its commitment to supply 20,000 Gravity SUVs to Nuro, which will then install its proprietary self-driving software and sensor suite before delivering them for Uber’s operations.

While the automotive sector currently represents a modest segment, accounting for less than 2 percent of Nvidia’s revenue in the last fiscal year, the world’s most valuable company foresees immense growth. This expansion is driven by the rise of "physical AI," which is poised to power a new generation of robots, autonomous vehicles, and automated factories. Should the market materialize and Uber successfully scale its self-driving fleet to 100,000 vehicles, they would significantly augment the nearly 9 million global human drivers presently engaged in ferrying paying passengers. Nvidia has already cultivated an extensive network of automotive partners, including Mercedes-Benz, JLR, and Volvo Cars, which are also adopting the Hyperion platform. Further demonstrating its commitment to a self-driving future, Nvidia partners with a diverse array of companies focused on solving autonomous driving challenges, such as Nuro, Pony.ai, Wayve, May Mobility, WeRide, and Avride, alongside driverless trucking innovators like Aurora, Volvo Autonomous Solutions, and Waabi.

At the core of these advancements lies the Nvidia Drive AGX Hyperion 10 architecture. This sophisticated platform comprises two Drive AGX Thor computers, the robust DriveOS operating system, and an approved board design. Nvidia has also validated a comprehensive sensor suite tailored for autonomous operation, including 14 cameras, nine radars, one lidar, and 12 ultrasonic sensors. However, the system offers flexibility, allowing customers to customize the sensor configuration to meet their specific requirements. For the self-driving software stack, partners have the option to build upon Nvidia’s proprietary Drive AV software or integrate a third-party solution, offering adaptability and choice in development.