Switch language한국어
Back to the list

Nvidia bets big on physical AI at GTC Taipei with a new world model, driving brain, and open humanoid robot

TL;DR AI

Key summary

2 min read
  1. Nvidia unveiled Cosmos 3, a multimodal world model for generating and interpreting training data across robotics, vehicles, and video systems.

  2. It also introduced Alpamayo 2 Super, a larger open driving model for Level 4 autonomy with trajectory output, meta-actions, and reasoning traces.

  3. Nvidia expanded its physical AI stack with an open humanoid robot reference platform, the Cosmos Coalition, AlpaGym, and OmniDreams.

  4. The moves are meant to make synthetic data, simulation, and reasoning more useful for real-world robotics and autonomous driving deployment.

Read the original