What’s new in Isaac Lab 3.0?
Train and deploy robot policies with a lightweight, modular framework built for flexible simulation and learning at scale.
Here are a few highlights:
🤖 Newton physics: Combine rigid-body, deformable, and particle simulation for dexterous manipulation.
🛠️ Flexible backends: Switch supported physics, rendering, and visualization without rewriting task logic.
👁️ Rendering options: Choose Newton Warp’s high throughput or OVRTX’s photorealistic visuals.
⚡ Training at scale: Scale parallel robot learning from one GPU to clusters.
🧠 Experimental RL post-training: Refine models like Isaac GR00T through simulated practice.
🚀 LEAPP deployment: Export portable policy bundles for deployment on real robots.
Read the announcement | Watch the Livestream
Stay up to date on the latest NVIDIA Robotics announcements by following us on LinkedIn, X, Instagram, Facebook, and Discord.