Key Highlights
- Nvidia revealed autonomous vehicle collaborations with BYD, Geely, Hyundai, Nissan, and Isuzu during its GTC conference
- Each partner plans to utilize Nvidia’s DRIVE Hyperion platform for developing Level 4 autonomous capabilities
- The Uber collaboration extends to include robotaxi deployments in 28 cities spanning four continents by 2028
- Initial deployment will begin in Los Angeles and San Francisco Bay Area during the first half of 2027
- Alpamayo 1.5 debuts as an enhanced open AI model for autonomous driving, with over 100,000 developer downloads
Nvidia’s GTC conference on Monday showcased a wave of autonomous vehicle partnerships, strengthening the chipmaker’s position in the self-driving sector. CEO Jensen Huang announced BYD, Geely, Hyundai, Nissan, and Isuzu as new collaborators for the DRIVE Hyperion platform during his presentation in San Jose.
DRIVE Hyperion represents Nvidia’s comprehensive AV platform. The solution integrates data center training, large-scale simulation capabilities, and in-vehicle computing within a unified reference architecture. Automakers can leverage this framework to create Level 4-capable vehicles — automobiles capable of autonomous operation without human intervention under specific conditions.
Huang expressed confidence in the technology’s maturity. “We’ve been working on self-driving cars for a long time. The ChatGPT moment of self-driving cars has arrived,” he told the audience.
The Uber partnership generated significant interest. Nvidia and Uber revealed an expanded agreement to establish fully autonomous vehicle fleets across 28 cities on four continents by 2028. Initial operations will launch in Los Angeles and the San Francisco Bay Area during the first half of 2027.
This fleet will operate using Nvidia’s complete AV software stack, incorporating the DRIVE Hyperion compute architecture alongside the newly introduced Halos OS safety system.
Bolt, Grab, and Lyft joined as additional mobility providers adopting DRIVE Hyperion, extending Nvidia’s AV platform reach beyond traditional automotive manufacturers.
Alpamayo 1.5 Advances AV Artificial Intelligence
Nvidia launched Alpamayo 1.5 on Monday, delivering substantial improvements to its open portfolio of AI models for autonomous driving. This updated version processes driving video, motion history, navigation guidance, and natural language prompts as inputs, generating driving trajectories with reasoning traces as outputs.
Developers can now use text prompts to instruct the car’s actions and behavior patterns. This represents an advancement over previous systems where modifying driving behavior demanded model retraining.
The initial Alpamayo model has reached more than 100,000 automotive developers through downloads since launching earlier this year. Version 1.5 introduces flexible multi-camera support and configurable camera parameters, simplifying AI stack reuse across various vehicle models.
Safety Architecture and Simulation Capabilities
Alongside partner announcements and model enhancements, Nvidia presented NVIDIA Halos OS — a unified safety architecture built upon ASIL D-certified foundations. This system provides AV developers with a production-ready safety layer for Level 4 vehicles.
Ten companies, including AEye, Hesai, Valeo, and Flex, joined the Nvidia Halos AI Systems Inspection Lab, established for testing and validating AV safety systems.
Nvidia also released NVIDIA Omniverse NuRec for general availability. NuRec employs 3D Gaussian Splatting to reconstruct real-world environments for simulation purposes, enabling developers to stress-test AV behavior without constructing physical test tracks.
Isuzu and TIER IV are utilizing DRIVE Hyperion for developing Level 4 autonomous buses. Nissan’s L4 program utilizes Wayve software operating on the platform.
Nvidia stock climbed 0.26% during after-hours trading on Monday, building upon regular session gains.
