Physical AI Accelerated by Three NVIDIA Computers for Robot Training, Simulation and Inference

From NVIDIA: 2024-10-24 02:25:13

ChatGPT ushers in the era of generative AI, changing digital work across industries. However, physical AI in industries like transportation and manufacturing has yet to excel. NVIDIA introduces multimodal, physical AI that can interact with, navigate, and understand the physical world through advanced training and simulation on three computers.

The rise of neural networks and GPUs has revolutionized AI, leading to Software 2.0 where software writes software. Multimodal and diffusion models in generative AI are advancing rapidly, enabling robots to navigate and interact in a 3D world. Humanoids robots, with a booming market projection, are set to transform industries worldwide.

NVIDIA’s three computers accelerate humanoid robot development by training foundation and generative AI models on a supercomputer with NeMo. Isaac Sim and Omniverse aid in simulation and validation of robot models, saving costs and ensuring performance. Robotic facilities utilizing synthetic data and digital twins promise seamless integration and operations.

The developer ecosystem benefits from NVIDIA’s technology, empowering companies like Universal Robots and RGo Robotics to create AI-enhanced robot applications. Humanoid robot makers like Boston Dynamics and Fourier are using Isaac Sim and Isaac Lab to build robots for varied industries, while Field AI develops multi-purpose foundation models for outdoor environments.



Read more at NVIDIA: Physical AI Accelerated by Three NVIDIA Computers for Robot Training, Simulation and Inference