Nvidia Unveils Omniverse Cloud APIs, Advancing Digital Twin Technology
Nvidia has launched its Omniverse Cloud application programming interfaces (APIs), marking a significant advancement in digital twin software tools.
With key industry players like Ansys, Cadence, and Siemens participating, these APIs are poised to drive a new era of digital twin applications, enhancing design, simulation, and operational processes. This announcement was made at the Nvidia GTC 2024 event in San Jose, California.
The introduction of Omniverse Cloud APIs is a game-changer for digital twin technology. By integrating core Omniverse functionalities directly into existing software applications, developers can now refine workflows and speed up the creation of digital twins across various industrial sectors. I will be moderating a panel on industrial digitalization and digital twins on Tuesday at 4 p.m. during GTC.
“We envision a future where everything manufactured will have digital twins,” said Jensen Huang, CEO of Nvidia, during his keynote at Nvidia GTC 2024. “Omniverse serves as the operating system for creating and managing physically realistic digital twins. By leveraging generative AI and Omniverse, we are revolutionizing heavy industries, unlocking limitless opportunities for creators.”
While progress in the consumer metaverse has been gradual, the industrial metaverse is gaining rapid momentum. Nvidia has devoted five years to developing Omniverse, utilizing Universal Scene Description (USD)—a data exchange format designed by Pixar for 3D simulations. For enterprises, Nvidia licenses Omniverse for $4,500 annually, and the introduction of Omniverse Cloud APIs is a highlight of this GTC event.
“Omniverse is a platform for building and deploying physically based industrial digitalization applications. It represents a feedback loop for AI to transition into the physical realm. Omniverse is not just a tool; it is an ecosystem that enhances existing tools, connecting to the world’s largest design, CAD, and simulation platforms such as Siemens, Autodesk, and Adobe,” stated Rev Lebaredian, Nvidia’s vice president of simulation, during a press briefing with GamesBeat. “Our software partners like Siemens and Ansys leverage Omniverse to create innovative tools and services. Customers like BMW, Amazon Robotics, and Samsung utilize Omniverse for developing their own tools, linking their industrial operations.”
Both generative AI and digital twins are transforming how companies across various sectors design, manufacture, and operate their products.
Traditionally, enterprises depended on physical prototypes and costly alterations for extensive industrial projects, a process that was expensive and prone to errors. This limited innovation and extended time-to-market, according to Lebaredian.
Digital twin theory suggests constructing a factory digitally to simulate all operations realistically, enabling iterative design refinement before breaking ground. Once operational, the digital twin facilitates quick reconfigurations based on real-time data gathered by sensors, optimizing production processes and significantly reducing costs.
Lebaredian noted that heavy industries are grappling with labor shortages, supply chain constraints, and a complex geopolitical landscape, driving the urgency to embrace autonomy and software solutions.
“They are creating digital twins to design, simulate, build, operate, and optimize assets and processes,” he explained. “This approach enhances operational efficiency, saving billions in costs. Once established, these digital twins become the foundation for robotic systems and virtual training environments for AI, ensuring that algorithms developed in virtual settings perform effectively in the real world.”
Omniverse Cloud APIs
The Nvidia Omniverse Cloud APIs will elevate the capabilities of the industrial metaverse.
The five newly launched Omniverse Cloud APIs offer a robust toolkit for developers, including:
1. USD Render: Generates fully ray-traced renders from OpenUSD data.
2. USD Write: Allows users to modify and interact with OpenUSD data.
3. USD Query: Facilitates scene queries and interactive scenarios.
4. USD Notify: Tracks changes in USD and provides updates.
5. Omniverse Channel: Enhances collaboration by connecting users and tools across scenes.
“Omniverse Cloud APIs empower developers to integrate Omniverse technologies for OpenUSD and RTX directly into their existing applications and workflows,” said Lebaredian. “This provides ISVs (independent software vendors) with comprehensive Omniverse interoperability across real-time rendering tools and the potential for collaboration across devices. These APIs will be accessible on Microsoft Azure later this year, allowing developers to offer managed services for enterprises or consumers. The world’s leading industrial software companies are adopting these APIs, reaching hundreds of thousands of users, starting with Siemens.”
Companies such as Siemens and Ansys are already incorporating Omniverse Cloud APIs into their platforms. Siemens, for instance, is enhancing its Xcelerator Platform by integrating these APIs into its cloud-based product lifecycle management software, Teamcenter X.
“Through the Nvidia Omniverse API, Siemens equips customers with generative AI to create even more immersive physics-based digital twins,” stated Roland Busch, president and CEO of Siemens AG. “This integration enables teams to design, build, and test next-generation products and manufacturing processes virtually before physical production begins.”
I interviewed Busch regarding digital twins at CES 2024 in January.
Ansys and Cadence are also utilizing Omniverse Cloud APIs to ensure data interoperability and real-time visualization within their software solutions for autonomous vehicles and digital twin platforms.
As the demand for autonomous machines rises, the need for efficient workflows becomes crucial. Omniverse Cloud APIs bridge the gap between simulation tools, sensor solutions, and AI-based monitoring systems, facilitating the rapid development of robots and autonomous vehicles.
“The next phase of industrial digitalization has begun,” emphasized Andy Pratt, corporate vice president at Microsoft Emerging Technologies. “With Nvidia Omniverse APIs hosted on Microsoft Azure, organizations worldwide can connect, collaborate, and enhance their existing tools to pioneer the next generation of AI-enabled digital twins.”
The launch of Omniverse Cloud APIs represents a transformative moment in the digital transformation of industries globally. By harnessing Nvidia’s AI-driven technology, companies can achieve unprecedented innovation and efficiency, revolutionizing their design, simulation, building, and operational processes in the digital era.
Industrial software companies utilizing Omniverse Cloud APIs include Ansys, Cadence, Dassault Systèmes, Hexagon, Microsoft, Rockwell Automation, Siemens, and Trimble.
Nvidia and Siemens Partnership
Siemens is fully committed to embracing Omniverse and digital twin technology. The Siemens Xcelerator platform provides immersive visualization tools that enhance product lifecycle management. Siemens has begun adopting the new Nvidia APIs within its Xcelerator platform applications, starting with Teamcenter X.
Teamcenter X is Siemens’ cloud-based product lifecycle management (PLM) software relied upon by enterprises of all sizes to scale product development.
By integrating Nvidia Omniverse with Teamcenter X, Siemens equips its engineering teams to create highly immersive and photorealistic physics-based digital twins, minimizing workflow inefficiencies and reducing errors.
In collaboration with Sony, Siemens demonstrated how enhanced XR design can engage users more effectively.
Using Omniverse APIs, workflows such as material application, lighting environments, and supporting scenery assets in physically based renderings will be significantly accelerated through generative AI.
AI integrations will contextualize engineering data to reflect real-world conditions, providing stakeholders—including sales and marketing teams, decision-makers, and customers—with enhanced insights into product appearances.
By linking Omniverse Cloud APIs with the Xcelerator platform, Siemens allows customers to improve their digital twins with physically based rendering, propelling industrial-scale design and manufacturing projects forward. With generative AI capabilities, users can seamlessly generate 3D objects or high-dynamic range image backgrounds to view assets in context.
This integration empowers companies like HD Hyundai, a leader in sustainable ship manufacturing, to visualize complex engineering challenges directly within Teamcenter X. At Nvidia GTC, Siemens and Nvidia showcased how HD Hyundai can use the software to visualize digital twins of liquefied natural gas carriers, which may consist of over 7 million parts, aiding product validation before moving into production.
Revolutionizing Data Center Design
Nvidia leverages Omniverse to design digital twins for advanced data centers.
The company also introduced a digital blueprint for constructing next-generation data centers using Omniverse digital twins, supported by Ansys, Cadence, Patch Manager, Schneider Electric, Vertiv, and others.
Nvidia noted that developing, simulating, and launching modern data centers is a complex task requiring careful consideration of performance, energy efficiency, and scalability. It involves collaboration among teams of highly skilled engineers across various fields, including compute/network design, CAD modeling, and mechanical, electrical, and thermal engineering. These inevitably refer to the "sovereign AI" data centers Jensen Huang has frequently discussed.
On the exhibition floor, Nvidia showcased a fully operational digital twin of a data center designed within Omniverse. The innovative design includes AI supercomputers within a liquid-cooled system, featuring two racks, 18 Grace CPUs, and 36 Nvidia Blackwell GPUs, all interconnected via fourth-generation Nvidia NVLink switches.
To expedite the launch of new data centers, Nvidia initially created its digital twin using software tools connected through Omniverse. Engineers unified and visualized multiple CAD datasets with complete physical and photorealistic accuracy in Universal Scene Description (OpenUSD), utilizing the Cadence Reality digital twin platform powered by Omniverse APIs.
The newly implemented GB200 cluster is set to replace an existing cluster in one of Nvidia’s legacy data centers. To initiate the digital build, Kinetic Vision employed the NavVis VLX wearable lidar scanner to generate precise point cloud data and panoramic images of the facility.
Using Prevu3D software, engineers removed the current clusters and converted the point cloud into a 3D mesh, producing an accurate 3D model of the facility to simulate the new digital data center.
By leveraging the Cadence Reality platform integrated with Omniverse, engineers visualized multiple CAD datasets with enhanced precision and realism. The integration offered a robust computing environment that enabled teams to develop OpenUSD-based 3D applications seamlessly.
Omniverse Cloud APIs also enhanced interoperability with various tools, including Patch Manager and Nvidia Air. Utilizing Patch Manager, the engineering team designed the physical layout of their cluster and networking infrastructure, ensuring accurate cabling lengths and proper routing.
The demo illustrated how digital twins can enable users to thoroughly test, optimize, and validate data center designs before their physical counterparts are produced. By visualizing data center performance through digital twins, teams can fine-tune designs and prepare for potential scenarios.
Moreover, users can refine data center and cluster designs by balancing various boundary conditions—such as cabling lengths, power, cooling, and space—integrated into the overall design process. This allows engineers and design teams to bring their clusters online more efficiently and effectively than ever before.
Ranking
