.svg){kind=spacer}
Details
- NVIDIA has made the open-source Newton Physics Engine available within Isaac Lab, advancing robotic simulation technology after its co-development with Google DeepMind and Disney Research earlier this year.
- The update also introduces the open NVIDIA Isaac GR00T N1.6 model, which enhances robots' skill acquisition by fusing reasoning, vision, language, and action capabilities.
- Newton offers GPU-accelerated, differentiable physics simulation, narrowing the gap between simulated and real-world robotic environments.
- Isaac Lab provides a unified environment for robot learning, supporting varied workflows from reinforcement learning to imitation learning, and includes a wide selection of pre-built environments and robot models.
- This integration allows researchers and developers to pair Newton’s advanced physical modeling with Isaac Lab’s training tools, supporting tasks including dexterous manipulation, coordinated multi-agent learning, and whole-body control for large-scale robotics research.
Impact
NVIDIA’s move reinforces its leadership in robotic AI platforms, challenging established simulation engines with a GPU-native approach and open-source philosophy. By merging high-fidelity simulation and cutting-edge AI skill models, researchers can accelerate development and deployment of advanced robotics across industries such as manufacturing, healthcare, and automation.
