Uber and Nvidia announced the expansion of their autonomous driving partnership, with plans to launch Nvidia's full stack software driven autonomous driving fleets in Los Angeles and San Francisco in the first half of 2027, and expand to 28 cities worldwide by 2028.
The core of this deployment is NVIDIA's DRIVE Hyperion autonomous driving platform and the new generation of inference based AI model "Alpamayo". This model is designed to address complex "long tail" scenarios such as unpredictable construction areas or pedestrian behavior, marking Nvidia's evolution towards a full stack L4 level (highly autonomous driving) software provider.