NVIDIA Announces Release of Omniverse With New USD Connectors and Tools, Simulation Technologies, and Developer Frameworks
August 11, 2022
NVIDIA announced a new range of developer frameworks, tools, apps, and plugins for NVIDIA Omniverse, the platform designed for building and connecting metaverse worlds based on Universal Scene Description (USD).
The expansion of Omniverse includes several AI-powered tools and features that deliver artists, developers, and engineers the power to build virtual worlds and content, and connect to today’s 3D applications including PTC Creo, SideFX Houdini, Unity, and solutions from the Siemens Xcelerator platform.
According to the company, Omniverse is already in use by some 700 companies globally to enhance architectural and product design, simplify visual effects workflows, and build digital twins of factories, cities, and the planet.
In addition to launching the NVIDIA Omniverse Avatar Cloud Engine — a suite of cloud-native AI models and services for building and deploying lifelike virtual assistants and digital humans — the company unveiled the following platform updates:
- Omniverse Kit — a toolkit for building native Omniverse extensions and applications which includes updates to PhysX® in Omniverse, including real time, multi-GPU scalable soft body and particle cloth simulation, helping bring more physical accuracy to virtual worlds and objects, and new OmniLive Workflows for an an overhaul of USD-based collaboration in Omniverse, bringing increased speed and performance to multiple app 3D workflows, and enabling non-destructive USD workflows to make collaboration between artists and developers ideal.
- Omniverse Audio2Face — an AI tool that can create facial animations directly from an audio file introducing AI-driven animation with the new ability to infer and generate realistic emotions and animates all facial features.
- Omniverse Machinima — an app for building 3D cinematics and animated films launches hundreds of new, free 3D assets from Post Scriptum, Beyond the Wire, and Shadow Warrior 3 games, plus, a suite of new AI animation tools like Audio2Gesture - an AI that generates realistic arm and body motion from an audio file.
- Omniverse DeepSearch — now available for Omniverse Enterprise customers, DeepSearch helps teams use AI to intuitively and accurately search through untagged 3D asset databases of visuals using natural language like “red, rusty barrel.” DeepSearch is a game changing tool for game developers or VFX studios who have hundreds of thousands of untagged assets that pose an immense challenge to search through. The Industrial Light & Magic is leveraging DeepSearch to unlock its library of environment assets.
Additionally, NVIDIA Modulus, a physics machine-learning framework, is newly available as an Omniverse Extension. Per the company, Modulus-trained Physics ML models are 4,000x and even 100,000x faster depending on the application, while providing accuracy closer to high-fidelity simulations. Modulus is one of the cornerstones of scientific digital twins, including NVIDIA’s Earth-2.
As part of a collaborative effort with its partners in the industrial, design, simulation, and CAD software ecosystems, NVIDIA also unveiled 11 new Omniverse Connectors, which are USD-based plugins, further opening Omniverse workflows to companies in the industrial and scientific communities. These bring the total number of Connectors to the Omniverse USD ecosystem to 112.
Newly available in beta are Connectors for PTC Creo, Visual Components, and SideFX Houdini. NVIDIA announced the ongoing development of Connectors for Blender, Autodesk Alias and Autodesk Civil 3D, Siemens JT, SimScale, Open Geospatial Consortium, and Unity, which will further unlock metaverse workflows for manufacturing, engineering, and design companies.
NVIDIA also released updates to the core simulation technologies that represent materials, physics, and light in metaverse worlds.
- NVIDIA MDL — which has served for 10 years as the material standard for physically accurate representation of 3D materials — is now fully open sourced, enabling developers to bring material definition language support to any renderer.
- NeuralVDB, coming soon to beta, is the next evolution of OpenVDB and brings AI and GPU optimization to sparse volume datasets and reduces memory footprint of datasets by up to 100x.
For more information, visit: https://www.nvidia.com/en-us/omniverse/