NVIDIA Unveils GR00T N1 and Newton at GTC 2025, Paving the Way for Advanced Humanoid Robotics​.

At the 2025 NVIDIA GTC conference in San Jose, CEO Jensen Huang captivated an audience of approximately 16,000 with a keynote that spotlighted significant advancements in robotics. Central to his presentation were the introductions of GR00T N1, an open-source foundation model for humanoid robots, and Newton, a collaborative physics engine designed to enhance robot learning and development.​

GR00T N1 is engineered to provide a versatile AI framework capable of perception, reasoning, and action across diverse scenarios. Developers can fine-tune this model for specific applications, leveraging its synthetic data generation capabilities to enrich training datasets. By open-sourcing GR00T N1 on platforms like Hugging Face, NVIDIA aims to lower barriers for researchers and developers, fostering innovation in the field of humanoid robotics.​

Complementing GR00T N1, NVIDIA introduced Newton, an open-source, extensible physics engine developed in collaboration with Google DeepMind and Disney Research. Newton is compatible with MuJoCo, a widely used physics engine in robotics research, and is designed to simulate complex dynamics and contact-rich environments. This tool aims to provide more accurate predictions of robot interactions with their environments, enhancing the development of dexterous manipulation and locomotion capabilities.​

In his keynote, Huang also highlighted NVIDIA's partnership with General Motors to develop next-generation vehicles, automotive factories, and robots. This collaboration underscores NVIDIA's commitment to integrating AI across various industries, including automotive manufacturing.​

Further emphasizing the company's dedication to advancing robotics, NVIDIA announced the Cosmos platform, a suite of generative world foundation models, advanced tokenizers, guardrails, and an accelerated data processing pipeline. Cosmos is designed to power world model training and accelerate physical AI development for autonomous vehicles and robots.​

Agility Robotics and Boston Dynamics also revealed enhancements to their humanoid robots through collaborations with NVIDIA. Agility's Digit robot now incorporates an NVIDIA accelerator for onboard AI models, utilizing IsaacSim and IsaacLab for cloud-based training. Boston Dynamics' Atlas robot benefits from the NVIDIA Jetson Thor computing platform, enabling the execution of complex, multimodal AI models in conjunction with whole-body and manipulation controllers.​

The conference featured a session by Dr. Ken Goldberg titled "Igniting the Real Robot Revolution Requires Closing the Data Gap," where he discussed the evolution of AI and robotics over the past five years, emphasizing the importance of data in advancing the field.​

GTC 2025 marked a significant milestone in NVIDIA's journey to revolutionize robotics, showcasing the company's commitment to developing open-source tools and fostering collaborations that drive innovation in AI and humanoid robot development.​

Source: https://www.therobotreport.com/the-biggest-robotics-news-from-nvidia-gtc-day-2/

This is non-financial/medical advice and made using AI so could be wrong.