NVIDIA’s Physical AI Stack Is Starting to Look Like the AWS of Robotics
This image was generated by AI and may not depict real events.
NVIDIA is developing a comprehensive robotics platform, dubbed its 'Physical AI Stack', to simplify robot development and deployment. The platform includes several interconnected layers, such as GR00T, Cosmos, Isaac Sim, and Newton, to facilitate a seamless sim-to-real workflow.
NVIDIA is shifting its focus in robotics from showcasing individual robots to developing a comprehensive platform to support robot development. The company's 'Physical AI Stack' includes several layers: GR00T for robot reasoning and action, Cosmos for generating synthetic training data, Isaac Sim and Isaac Lab for simulation and testing, and Newton for improving physics simulations. These layers work together to create a sim-to-real loop, allowing robots to learn and test in synthetic environments before being deployed in the real world. This approach is likened to Amazon Web Services (AWS) in cloud computing, providing a default platform for many teams to build upon. By providing a connected workflow, NVIDIA aims to make robotics development faster, more efficient, and more reliable. The platform is designed to address the harder problems in robotics, such as handling messy, expensive, and unpredictable environments.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.