\n\tNew NVIDIA\u00a0Omniverse\u00a0NuRec\u00a03D Gaussian Splatting Libraries Enable Large-Scale World Reconstruction
\n\tNew NVIDIA Cosmos\u00a0Models\u00a0Enable World Generation and Spatial Reasoning
\n\tNew NVIDIA\u00a0RTX\u00a0PRO Blackwell Servers and NVIDIA DGX Cloud Let Developers Run the Most Demanding Simulations Anywhere
\n\tPhysical AI Leaders Amazon Devices & Services, Boston Dynamics, Figure AI and Hexagon Embrace Simulation and Synthetic Data Generation
<\/p>\n
SIGGRAPH\u2014NVIDIA today announced new NVIDIA Omniverse<\/a>\u2122 libraries and NVIDIA Cosmos<\/a>\u2122 world foundation models (WFMs<\/a>) that accelerate the development and deployment of robotics solutions.<\/p>\n Powered by new NVIDIA RTX PRO\u2122 Servers<\/a> and NVIDIA DGX\u2122 Cloud<\/a>, the libraries and models let developers anywhere develop physically accurate digital twins<\/a>, capture and reconstruct the real world in simulation, generate synthetic data<\/a> for training physical AI<\/a> models and build AI agents that understand the physical world.<\/p>\n \u201cComputer graphics and AI are converging to fundamentally transform robotics,\u201d said Rev Lebaredian, vice president of Omniverse and simulation technologies at NVIDIA. \u201cBy combining AI reasoning with scalable, physically accurate simulation, we\u2019re enabling developers to build tomorrow\u2019s robots and autonomous vehicles that will transform trillions of dollars in industries.\u201d<\/p>\n New NVIDIA Omniverse Libraries Advance Applications for World Composition Omniverse NuRec rendering is now integrated in CARLA<\/a>, a leading open-source simulator used by over 150,000 developers. Autonomous vehicle<\/a> (AV) toolchain leader\u00a0Foretellix<\/a> is integrating NuRec, NVIDIA Omniverse Sensor RTX<\/a>\u2122 and Cosmos Transfer to enhance its scalable synthetic data generation with physically accurate scenarios. Voxel51\u2019s<\/a> data engine for visual and multimodal AI, FiftyOne, supports NuRec to ease data preparation for reconstructions. FiftyOne is used by customers such as Ford and Porsche.<\/p>\n Boston Dynamics, Figure AI, Hexagon<\/a>, RAI Institute, Lightwheel<\/a> and Skild AI are adopting Omniverse libraries, Isaac Sim and Isaac Lab<\/a> to accelerate their AI robotics development, while Amazon Devices & Services<\/a> is using them to power a new manufacturing solution.<\/p>\n Cosmos Advances World Generation for Robotics New models<\/a> announced at SIGGRAPH deliver major advances in synthetic data generation speed, accuracy, language support and control:<\/p>\n
Lightwheel, Moon Surgical and Skild AI are using Cosmos Transfer to accelerate physical AI training by simulating diverse conditions at scale.<\/p>\n Cosmos Reason Breaks Through World Understanding NVIDIA Cosmos Reason<\/a> \u2014 a new open, customizable, 7-billion-parameter reasoning VLM for physical AI and robotics \u2014 lets robots and vision AI agents reason like humans, using prior knowledge, physics understanding and common sense to understand and act in the real world.\u00a0<\/p>\n Cosmos Reason can be used for robotics and physical AI applications including:<\/p>\n NVIDIA\u2019s robotics and NVIDIA DRIVE<\/a>\u2122 teams are using Cosmos Reason for data curation and filtering, annotation and VLA post-training. Uber is using it to annotate and caption AV training data.<\/p>\n Magna<\/a>\u00a0is developing with Cosmos Reason as part of its City Delivery platform \u2014 a fully autonomous, low-cost solution for instant delivery \u2014 to help vehicles adapt more quickly to new cities. Cosmos Reason adds world understanding to the vehicles\u2019 long-term trajectory planner. VAST Data<\/a>, Milestone Systems and\u00a0Linker Vision<\/a> are adopting Cosmos Reason to automate traffic monitoring, improve safety and enhance visual inspection in cities and industrial settings.<\/p>\n New NVIDIA AI Infrastructure Powers Robotics Workloads Anywhere Accelerating the Developer Ecosystem
\nNew NVIDIA Omniverse software development kits (SDKs) and libraries<\/a> are now available<\/a> for building and deploying industrial AI<\/a> and robotics simulation<\/a> applications.<\/p>\n
\nCosmos WFMs, downloaded over 2 million times, let developers generate diverse data for training robots at scale using text, image and video prompts.<\/p>\n
\n\tCosmos Transfer-2<\/a>, coming soon, simplifies prompting and accelerates photorealistic synthetic data generation from ground-truth 3D simulation scenes or spatial control inputs like depth, segmentation, edges and high-definition maps.
\n\tA distilled version of Cosmos Transfer reduces the 70-step distillation process to one so developers can run the model on NVIDIA RTX PRO Servers<\/a> at unprecedented speed.
<\/p>\n
\nSince the introduction of OpenAI\u2019s CLIP model, vision language models<\/a> (VLMs) have transformed computer-vision tasks like object and pattern recognition. However, they have not yet been able to solve multistep tasks nor handle ambiguity or novel experiences.<\/p>\n
\nTo enable developers to take full advantage of these advanced technologies and software libraries, NVIDIA announced AI infrastructure designed for the most demanding workloads.<\/p>\n
\nTo help robotics and physical AI developers advance 3D and simulation technology adoption, NVIDIA also announced:<\/p>\n
\n\tOpenUSD Curriculum and Certification, which addresses demand for USD expertise, with support from\u00a0AOUSD<\/a> members Adobe, Amazon Robotics, Ansys \u2014 part of Synopsys, Autodesk, Pixar, PTC, Rockwell Automation, SideFX, Siemens, TCS and Trimble, as well as industry leaders such as Hexagon.
\n\tOpen-source collaboration with Lightwheel to integrate robot policy training and evaluation frameworks into NVIDIA Isaac Lab, featuring parallel reinforcement learning training capabilities, benchmarks and simulation-ready assets for robot manipulation and locomotion.
<\/p>\n