US EditionNvidia Unveils Self-Driving Car Tech as Part of Physical AI Push
Nvidia's Alpamayo includes a 10-billion-parameter model and over 1,700 hours of driving data to improve autonomous vehicle safety and decision transparency.
- On Jan. 05, 2026, Nvidia unveiled the Alpamayo family at CES 2026, releasing its open-source AI models and code for free access on Hugging Face.
- To tackle long-tail scenarios, Nvidia says Alpamayo uses chain-of-thought reasoning to break problems into steps and select the safest path, Ali Kani said Monday.
- With a 10-billion-parameter architecture, Alpamayo 1 offers open model weights and inferencing scripts, and Nvidia releases Physical AI Open Datasets with 1,700+ hours of driving data plus AlpaSim and Cosmos for training.
- Mobility leaders including Lucid Motors, JLR, Uber and Berkeley DeepDrive are showing interest, while developers can fine-tune Alpamayo into runtime models and validate performance in simulation before deployment.
- Later this year, Nvidia's Rubin AI chips will release, and a driverless Mercedes‑Benz CLA will debut in the US in coming months; Huang said, `It drives so naturally because it learned directly from human demonstrators.
37 Articles
37 Articles
Nvidia Wants to Take Tesla’s Self-Driving Business
Nvidia Corp. (NASDAQ: NVDA) CEO Jensen Huang plans to jump into the self-driving car business with his “Alpamayo” AI platform. The world’s most valuable company commented at the launch of the system: “NVIDIA Alpamayo is an open portfolio of AI models, simulation frameworks, and physical AI datasets designed to accelerate the development of safe, transparent, and reasoning-based autonomous vehicles.” 24/7 Wall St. Key Points Nvidia Corp. (NASDA…
Nvidia announced on the ESC in Las Vegas the launch of the Alpamayyo, an AI technology with "reasonation" that helps autonomous machines process "the rare scenarios", drive in a complex environment and explain the driving decisions. The company has already started production of the Mercedes-Benz CLA model without driver, equipped with its platform, following the launch to take place in the US in the following months, then in Europe and Asia.
Coverage Details
Bias Distribution
- 67% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
