- NVIDIA rolls out Omniverse NuRec libraries and Cosmos Reason model powered by RTX PRO servers and DGX Cloud for immersive robotics simulation.
- Early adopters include Amazon, Boston Dynamics and Hexagon—highlighting momentum in AI‑driven physical simulation.
What happened: NVIDIA introduces next-gen Omniverse libraries and Cosmos models
At the SIGGRAPH conference, NVIDIA announced new Omniverse NuRec libraries and Cosmos world‑foundation models (WFMs) designed to accelerate robot development with physically accurate world reconstruction and synthetic data generation. These tools are powered by RTX PRO Blackwell Servers and DGX Cloud, enabling developers globally to build highly realistic digital twins and AI agents that navigate and understand physical environments.
Leading robotics firms—such as Amazon Devices & Services, Boston Dynamics, Figure AI and Hexagon—are already deploying these technologies, signalling strong enterprise demand for advanced simulation capabilities.
Also read: Solana developer claims quantum-resistant blockchain upgrade
Also read: Spirit Blockchain Capital targets yield generation from Dogecoin
Why it’s important
This announcement marks a pivotal shift in robotics development, embedding advanced AI reasoning with scalable, photo-realistic simulation. By integrating AI with OpenUSD-based Omniverse libraries, developers can now train models using rich, synthetic environments that closely mimic physical reality—accelerating robotics R&D while reducing dependence on costly real-world data.
Moreover, the adoption by industry leaders underscores a broader trend toward physical AI, where simulation fidelity and AI reasoning are increasingly merged. NVIDIA’s end-to-end ecosystem—from libraries to compute infrastructure—positions it at the centre of the robotics simulation renaissance, enabling faster innovation in autonomous systems, industrial automation and more.
