Digitalization is revolutionizing industries with the aim of breaking down traditional boundaries, accelerating technology convergence, and driving new industry convergence. The digital era is an era in which the physical and digital worlds merge. NVIDIA Omniverse™ is an enabler of this transformation, enabling digitalization by bridging the digital and physical worlds.
NVIDIA Omniverse :
Bridging the gap between AI and the digital age, connecting the virtual and the real
The world of 3D is ushering in a new era of design, simulation, and AI advancements. NVIDIA Omniverse is NVIDIA's easily scalable, open computing platform built for virtual collaboration and physically accurate, real-time simulations. It provides a shared virtual space for creators, designers, and engineers to collaborate seamlessly, and developers and enterprise technologists to customize and enhance applications based on Omniverse to accelerate workflows and projects. Since the launch of the Omniverse platform, NVIDIA has continued to update and iterate to enable it to more broadly empower enterprise innovation processes and digital transformation.
At GTC 2024, NVIDIA announced that NVIDlA Omniverse™ Cloud will be available as an API, extending the reach of this advanced industrial digital twin application and workflow creation platform to the entire ecosystem of software manufacturers. With five new Omniverse Cloud application programming interfaces (APIs), developers can easily integrate Omniverse's core technologies directly into existing digital twin design and automation software applications, or into simulation workflows for testing and validating autonomous machines such as robots or autonomous vehicles. At the same time, a new software framework built on Omniverse Cloud APIs enables developers to easily send their OpenUSD industrial scenes from content creation applications to the NVIDIA Graphics Delivery Network (GDN) and stream advanced 3D experiences to Apple Vision Pro.
Omniverse is already gaining enterprise-level adoption in several areas, including digital human creation, digital twin factories, autonomous driving, and more. It is not just a simple software tool, but a bridge between the physical and virtual worlds, a stage for creativity and technology.
The core functionality and features of the Omniverse platform include the following:
- Scalable visualization and simulation: With NVIDIA RTX™ and AI, Omniverse delivers photorealistic ray-traced effects to create full-fidelity product visualizations and simulation environments. Whether you're in game development, architectural visualization, or film and television production, you can benefit from the physically-accurate visuals created by Omniverse.
- Data Interoperability: In the Omniverse platform, lossless, interoperable workflows can be unlocked, reducing data transfer and import/export between disparate software and applications for seamless integration. This means that whether you're using Autodesk Maya, Blender, or any other program that supports Omniverse, you can work with each other on the Omniverse platform without worrying about compatibility.
- Modular Ecosystem and Easy-to-Use Development Tools: The Omniverse platform features a flexible, modular platform that allows users to quickly create and deploy custom workflows and applications based on low-code in graphics or Python without extensive operational programming knowledge.
- Built-in artificial intelligence: Predictive capabilities and natural language processing automate repetitive tasks. From simulation testing of self-driving cars to behavioral programming of robots, Omniverse provides a safe "testing ground."
OpenUSD: The 3D ecosystem on the way to the future
If Omniverse is the bridge, then the Universal Scene Description (OpenUSD) is the foundation, the foundation of Omniverse, an extensible framework and ecosystem for describing, compositing, simulating, and collaborating in 3D worlds.
Pixar introduced Universal Scene Description (USD) in 2012, an open API universal scene description tool for creating, editing, querying, rendering, collaborating, and simulating real-world scenes in virtual worlds.
USD was open-sourced in 2016, and due to its various powerful features, it is widely used not only in the field of visual effects, but also in architecture, engineering, design, retail, robotics, manufacturing, and other fields. More than just a file format, OpenUSD is an open, powerful, and easily extensible world composition framework with APIs for creating, editing, querying, rendering, mocking, and collaborating in virtual worlds.
▲ OpenUSD 示意图
It demonstrates its power as a large-scale industrial process technology
OpenUSD is an open standard thanks to four unique strengths: it is a powerful tool for 3D data modeling and resource exchange, which is highly scalable and meets the needs of virtual worlds.
Compositing engine
OpenUSD's efficient referencing and layering capabilities enable teams to make real-time, non-destructive edits to projects and preserve all source content, enabling iterations of collaborative workflows. This allows users to collaborate in real-time without interfering with each other through the system. For example, while one artist is modeling a scene, other artists can light the scene.
Custom mode
OpenUSD's data model can be extended with custom patterns, and OpenUSD already binds core patterns such as geometry and shading. NVIDIA has also partnered with Pixar and Apple to create patterns based on rigid body physics motion, and is also exploring other custom patterns to further expand the ecosystem of digital twins and virtual worlds.
Asset parser and data store
OpenUSD data is not tied to a file system or other persistent storage, and can even be generated programmatically. This is made possible by OpenUSD's plug-in system, which supports asset parsers like NVIDIA Omniverse™ Nuclues, as well as file formats like OBJ.
Hydra
Originally designed as an abstraction layer between the OpenUSD input scene graph and an output renderer such as RTX, Hydra has evolved into a general-purpose framework for working with composite scene graphs and runtimes. One of the main advantages of Hydra is that it is not tightly coupled to any single runtime data layout, which allows NVIDIA Omniverse to compile the combined OpenUSD into a deep vectorized data layout, known as a fabric. This means that if the renderer is adapted to Hydra, it will be able to support OpenUSD. With Hydra, OpenUSD can provide out-of-the-box services for different renderers such as Pixar Storm and RenderMan.
Based on OpenUSD's high degree of flexibility, scalability, collaboration, and customizability, which makes it possible to simulate the real world with high fidelity and accuracy, the 3D world of the future requires an HTML-like way to describe the individual interconnected applications in high detail. NVIDIA Omniverse is the platform for creating and operating metaverse applications, and OpenUSD is the foundation of NVIDIA Omniverse. NVIDIA Omniverse is built entirely on OpenUSD, an open and extensible ecosystem for compositing, simulating, and collaborating in 3D worlds.
Build workflows with Omniverse and OpenUSD
NVIDIA is working with partners and the open source community to ensure that USD is optimized and accessible to all, and NVIDIA has joined Pixar, Adobe, Apple, and Autodesk to form the OpenUSD Alliance (AOUSD) to promote the standardization, development, evolution, and growth of USD. With the addition of more than a dozen new members (Cesium, Chaos, Epic Games, Foundry, Hexagon, IKEA, Lowe's, Meta, OTOY, SideFX, Spatial, and Unity), OpenUSD's functionality and compatibility will be further enhanced, supporting not only the various needs of building virtual worlds, such as geometry, cameras, Lighting and materials are more realistic, and the full-stack layout of digital tools is complete, making it the central hub for 3D connectivity.
From creating digital twins of automotive factories to building custom scene manipulation tools, and advanced animation, OpenUSD and Omniverse are being used to accelerate a wide range of 3D workflows, including 3D model generation, 3D storytelling, AI-generated digital art, custom tool development, industrial digitalization, multi-application collaboration, and more.