NVIDIA Opens Portals to World of Robotics With New Omniverse Libraries, Cosmos Physical AI Models and AI Computing Infrastructure

Core Insights - NVIDIA has announced new Omniverse libraries and Cosmos world foundation models to enhance robotics development and deployment [1][11] - The integration of AI and computer graphics is set to revolutionize robotics, enabling the creation of advanced robots and autonomous vehicles [3][12] NVIDIA Omniverse Libraries - New software development kits (SDKs) and libraries are available for building industrial AI and robotics simulation applications [3] - Omniverse NuRec rendering is integrated into CARLA, a popular open-source simulator, enhancing synthetic data generation for autonomous vehicles [4] Adoption and Collaboration - Companies such as Boston Dynamics, Figure AI, and Amazon Devices & Services are adopting NVIDIA's Omniverse libraries to accelerate their AI robotics development [5] - Over 2 million downloads of Cosmos WFMs demonstrate significant interest in generating diverse training data for robots [6] Advances in Synthetic Data Generation - New models announced at SIGGRAPH improve the speed, accuracy, and control of synthetic data generation for robotics [6] - Cosmos Reason, a new vision language model, allows robots to reason and understand the physical world, enhancing their capabilities [8] AI Infrastructure - NVIDIA introduced AI infrastructure designed for demanding robotics workloads, including RTX PRO Blackwell Servers and DGX Cloud [10][12] - The infrastructure supports various development tasks, from training to simulation, across different platforms [13] Developer Ecosystem - NVIDIA is promoting the adoption of 3D and simulation technology through initiatives like OpenUSD Curriculum and Certification [17] - Collaboration with industry leaders aims to integrate advanced training frameworks into NVIDIA's robotics platforms [17]