Home Bots & Brains NVIDIA Unveils Open-Source Foundation Model for Humanoid Robots

NVIDIA Unveils Open-Source Foundation Model for Humanoid Robots

by Pieter Werner

NVIDIA has introduced the open-source Isaac GR00T N1 foundation model for humanoid robots at GTC 2025, aiming to enhance AI training and accelerate market adoption. According to TrendForce, GR00T N1 integrates extensive datasets, multi-modal inputs, and adaptable training capabilities, allowing researchers to fine-tune models using both real and synthetic data for various applications. The humanoid robotics market is projected to reach approximately $4 billion by 2028.

NVIDIA’s entry into this sector places it in competition with Figure AI and Google, which have developed their own humanoid robotics models, Helix and Gemini Robotics, respectively. These models incorporate rapid motion planning and cognitive decision-making to improve robotic autonomy. However, TrendForce suggests that NVIDIA’s dataset quality and deployment flexibility provide a competitive advantage.

NVIDIA CEO Jensen Huang has emphasized the role of humanoid robots in driving next-generation AI computing. The company has established a technology ecosystem supporting this development, offering embedded AI computing solutions such as Jetson Orin NX and Nano for mid-to-small humanoid robots, and AGX Orin for high-performance applications, with processing capabilities reaching up to 275 TOPS.

NVIDIA Omniverse is a key platform for training simulations and digital twins, widely used by Western humanoid robot developers as well as Chinese companies such as UBTECH, Unitree, and Fourier Intelligence. At GTC 2025, NVIDIA further expanded its robotics ecosystem with the introduction of the open-source physics engine Newton, the synthetic data generation tool GR00T Blueprint, and the GR00T N1 model.

While GR00T N1 is open-source, TrendForce notes that AI model parameters are often linked to specific hardware configurations. Developers utilizing GR00T N1 may adopt NVIDIA’s hardware solutions for optimized performance, reinforcing the company’s position in humanoid robotics. This approach mirrors NVIDIA’s strategy in the AI industry, where its CUDA technology and GPUs have contributed to its market leadership.

Misschien vind je deze berichten ook interessant

preload imagepreload image