Nvidia embraces European automakers

Nvidia is expanding its business in autonomous driving technologies across Europe. The ridesharing service Bolt relies on the AI company’s Alpamayo stack for its planned robotaxis, while VWN features the ID. Buzz AD with Nvidia Hyperion. Meanwhile, Volkswagen, Porsche, and other partners are advancing a shared data and simulation platform.
Nvidia CEO Jensen Huang at GTC Berlin
Image: Nvidia
The U.S. technology company Nvidia continues to strengthen its position in the autonomous driving market across Europe. At its AI conference GTC in Berlin, the world’s most valuable company by market capitalization at the time unveiled several partnerships and new technologies ranging from driver assistance systems to robotaxis and autonomous commercial vehicles. Among the companies involved are Mercedes-Benz, Volvo Cars, Volkswagen Nutzfahrzeuge, Lucid, Bolt, and Einride. Nvidia adopts a modular approach: its Hyperion vehicle platform is designed to work with both its own driving software and systems from other suppliers. Additionally, Nvidia provides tools for training AI models and virtually testing autonomous driving functions.
Bolt relies on Nvidia software for its robotaxis
The collaboration with mobility provider Bolt is particularly concrete. The company has chosen the Nvidia Alpamayo Stack as the foundation for its planned European robotaxi service. For this purpose, Lucid electric vehicles are set to be equipped with Nvidia’s Hyperion platform. Bolt and Nvidia aim to adapt the driving software for European roads and, in the future, integrate it into vehicles from other manufacturers as well. Bolt contributes its fleet management systems, operational experience, and local market knowledge.
The first test drives to develop and validate the technology are planned for next year in Madrid, the Spanish capital, subject to regulatory approvals. Bolt also specifically mentions Germany and France as key markets for deployment. Under this partnership, at least 25,000 Level-4 capable Lucid vehicles are set to be deployed in European markets. Bolt announced the partnership with Lucid back in September, initially planning Nvidia Hyperion as the technical platform without naming a specific driving software stack. With the now-announced selection of Alpamayo, the role distribution becomes clearer. Bolt aims to operate a total of 100,000 autonomous vehicles on its platform by 2035.
"Robotaxis represent the first application of robotics and physical AI for the mass market; they are revolutionizing the way people and goods move," said Nvidia CEO Jensen Huang at the event in Berlin. "Nvidia Hyperion provides our partners, along with our safety operating system Halos, with a comprehensive robotaxi platform designed for full autonomy and the highest level of safety. This brings us one step closer to a future where everything that is in motion moves autonomously."
Hyperion is intended to support various driving software solutions
The second important announcement concerns Volkswagen Commercial Vehicles. The company is showcasing the autonomous ID. Buzz AD based on Nvidia Hyperion. This platform combines computing hardware with a standardized sensor architecture and is designed to simplify the integration and validation of automated driving systems. Nvidia explicitly positions Hyperion as an open or modular foundation that can support not only its own Alpamayo Stack but also other driving software.
This is relevant in light of the current technology used in the ID. Buzz AD: The autonomous vehicle currently relies on Mobileye’s technology. Therefore, the introduction of the ID. Buzz AD with Hyperion does not necessarily mean that Volkswagen will replace Mobileye or has already decided on a new production architecture. Instead, Nvidia describes it as a platform that can integrate various autonomous driving systems. It remains unclear whether and in what specific configuration Hyperion will be used in future ID. Buzz AD models, and what role Mobileye will continue to play in this context.
This announcement should also be distinguished from Volkswagen’s search for a technology partner for autonomous driving in mass-market vehicles. As reported, VW is said to have chosen the British AI company Wayve after an internal comparison. According to “Spiegel,” Wayve will primarily provide the driving software that Cariad integrates into the existing vehicle architecture. This does not necessarily rule out Nvidia from the company’s technology ecosystem: Nvidia is invested in Wayve and supplies computing hardware for the startup’s autonomous driving projects. However, no specific combination of Wayve software and Nvidia hardware at Volkswagen has been confirmed yet.

Volvo and Mercedes Expand Cooperation
Volvo Cars also plans to rely on Hyperion and the Alpamayo Stack in the future. The Swedish manufacturer integrates Nvidia’s technologies with its own safety systems and vehicle development to advance autonomous driving in complex traffic situations. Mercedes-Benz, on the other hand, is deepening its partnership with Nvidia for Hyperion and Alpamayo to enhance its capabilities for Level 3 and Level 4 autonomous driving. Additionally, the jointly developed driver assistance system MB.Drive Assist Pro is set to receive approval for automated point-to-point driving in urban areas in Germany soon. However, it is “only” a driver assistance system, not a driverless Level 4 robotaxi.
Beyond passenger cars and robotaxis, Einride is also relying on Hyperion for the next generation of its autonomous driving system for electric trucks. The Swedish company plans to use this platform along with Nvidia’s Halos, Alpamayo, and Cosmos to develop and validate its Einride Driver. According to Einride, it remains responsible for the development and operation of the driving system. This illustrates another pattern in Nvidia’s strategy: the company increasingly provides a common technical foundation, while partners develop their own vehicles, applications, and business models.
Volkswagen relies on data and simulation
Another announcement relates to the development work behind driving functions. Volkswagen Group brands, Porsche, and Cariad are collaborating with Nvidia, Microsoft Azure, and the data platform provider Voxel51 to build a so-called “AV Data Flywheel.” Porsche is leading the initial implementation and integration of the involved systems. The goal is to systematically use data from real vehicle fleets for training, further development, and validation of automated driving systems.
Voxel51 handles data preparation and retrieval. Nvidia utilizes tools such as Cosmos Data Factory for processing and expanding training data, as well as Omniverse NuRec for reconstructing real driving scenarios in virtual environments. Microsoft Azure provides the cloud and computing infrastructure. The partners aim to selectively filter rare and safety-critical traffic situations from large datasets, replicate them in simulations, and test software versions in a reproducible manner. According to Nvidia, this approach could reduce the need for repeated physical testing by up to 90 percent and shorten development cycles by up to 85 percent. These are stated potential benefits, not yet proven results in actual operations.