NVIDIA Omniverse™ is a platform of APIs, SDKs, and services that enable developers to easily integrate Universal Scene Description (OpenUSD) and NVIDIA RTX™ rendering technologies into existing software tools and simulation workflows for building AI systems.
I'm excited to announce we'll be bringing together some of the brightest minds in technology and industry for my keynote at CES 2026 “Powering the Industrial AI Revolution”.
At the heart of it all: an on-stage conversation with Jensen Huang, Founder and CEO of NVIDIA. Together, we'll announce the next step in our partnership and how we can help companies around the world to adopt industrial AI at scale.
I'll also be joined by Jay Parikh from Microsoft, Dr. Athina Kanioura from PepsiCo, and Bob Mumgaard from Commonwealth Fusion Systems. Join us on site or online to learn more about how they are turning ideas into real-world impact with greater speed, quality, and efficiency.
Mark your calendars: CTA State of the Industry Address at 8:30 a.m., followed by my keynote at 9:00 a.m. (PST) on January 6th.
Looking forward to seeing you all in Las Vegas! 😉
Physical AI safety is entering a new phase as intelligent robots and robotaxis move from testing to deployment.🚦
With OpenUSD, NVIDIA Omniverse, and the NVIDIA Halos safety framework developers can prove how physical AI systems see, decide, and act before they ever touch the real world.
✔️ OpenUSD Core Specification 1.0 standardizes interoperable, simulation-ready assets across tools and partners.
✔️ SimReady assets, NVIDIA Isaac Sim, and Omniverse-powered digital twins enable safer, physics-accurate testing at scale.
✔️ NVIDIA Omniverse NuRec and Cosmos world foundation models turn real drives into long-tail edge cases for stress-testing safety.
Leaders like Bosch, Nuro, Wayve, and onsemi are already translating this into real-world safety gains.
🔗 Read the full blog https://nvda.ws/3L8qGPE
A huge milestone for OpenUSD. 👏
“The Core Specification is the definitive syntax for describing how virtual worlds are brought to life at scale.” - Aaron Luk
Read the newest release. 👇
Want to customize your OpenUSD training experience? Now you can. 🤝
✔️ Access the complete Learn OpenUSD curriculum on GitHub & adapt it to your 3D workflow or teaching style for your audience.
✔️Contribute improvements & localize the content to benefit the OpenUSD community, shaping the future of 3D development.
Get started: https://nvda.ws/4aX207b
Five years ago, during a quiet break, shortly after NVIDIA Omniverse was announced at GTC, I found myself revisiting Toy Story movies and casually exploring Pixar's RenderMan and their USD Kitchen Set, mostly out of curiosity. With early access to NVIDIA Omniverse, I started experimenting and ended up working on a short piece rendering Pixar’s Kitchen using OpenUSD, MDL materials, and real-time fluid simulation (video>).
https://lnkd.in/eA-CTraJ
At the time, Omniverse itself was very experimental, but OpenUSD was already surprisingly mature (Thanks to Pixar!).
Someone commented on my video:
“Is USD going to be the future? There’s a lot of development, but very few people are actually using USD in their workflow.”
I couldn’t really respond back then. Things were progressive, but it was all still early and uncertain.
Fast forward to 2025.
Omniverse has evolved into a serious digital twin platform, powered by the collaborative and integrative strengths of OpenUSD. With advances like the RTX 2.0 renderer, Index, realtime performance and visual fidelity have taken a big leap forward.
More importantly, we are now seeing real industrial adoption. Manufacturing, AEC, robotics, energy, and infrastructure teams are actively using OpenUSD-based pipelines for industrial digital twins.
NVIDIA has been supporting industrial adoption through the broader Omniverse platform and its growing set of openly available tools and reference implementations, many of which are published across its Omniverse GitHub repos. Alongside dedicated hardware platforms such as NVIDIA OVX and DGX architectures, there is also emerging cloud infrastructure like Brev.dev (Acquired by NVIDIA).
I was testing Brev recently, streaming full application experiences directly through the browser, including on Apple Silicon, which highlights how Omniverse is positioned not as a single piece of software, but as a scalable platform that can adapt to the size of the problem, the system complexity, and the fidelity of the data involved.
Another important signal of maturity is the work happening through the Alliance for OpenUSD (AOUSD). The working groups are actively shaping the evolution of OpenUSD across areas such as the core specification, geometry, materials, and physics, with participation from a broad range of industry and domain experts.
At community level, a growing number of independent groups and practitioners are working with this technology, which is evident from activity across Omniverse Discord servers and WhatsApp channels. The energy feels genuine, with communities such as the Omniverse India WhatsApp group now approaching nearly 1,000 active members in months.
At least from my point of view, that early question feels answered now: OpenUSD is not just “the future”. It is quietly becoming the backbone of how complex systems are designed, simulated, and understood.
#OpenUSD#NVIDIAOmniverse#DigitalTwins#Visualization#IndustrialAI#RTX#3D
Happy to share that I just earned the NVIDIA#OpenUSD Development Professional Certificate! 🎉
To tackle advanced end-to-end pipelines in robotics automation, I need an ecosystem that lets me build accurate digital twins, manage the composition of complex scenes, simulate robots and machines reliably, scale scenes with high performance, integrate training stacks for RL and VLA policies, and streamline deployment to production.
Not exactly easy requirements, right?
Luckily, that’s exactly what the NVIDIA Isaac ecosystem enables with #IsaacSIM, #IsaacLAB, and #IsaacROS
But the glue that ties it all together is OpenUSD: I can’t stress enough how valuable it is to have hands-on USD knowledge when working across the Isaac tools and keeping the sim-to-real pipeline flowing.
#OpenUSD#NVIDIA#Robotics#Automation#DigitalTwin#Simulation#ROS2
Take your AI career to the next level before the year ends.💡
For this month only, all NVIDIA Certification exams (Associate and Professional) are 50% off, giving you the chance to validate your knowledge in generative AI, OpenUSD, data science, and infrastructure & ops.
Take advantage of this one-time offer before it’s gone: https://nvda.ws/3KyeoQz
Real-world traffic data is essential, but it’s often sparse, expensive, or dangerous to collect—especially for critical edge cases.
We’ve added a new "recipe" to the NVIDIA Cosmos Cookbook that solves this by generating synthetic data for a broad range of traffic scenarios. It details an end-to-end workflow to:
1️⃣ Script complex, multi-agent scenarios in a virtual city environment.
2️⃣ Generate photorealistic video outputs using NVIDIA Cosmos.
3️⃣ Train perceptive VLMs for safer smart city applications.
Stop hunting for data. Start generating it.
Get the recipe: 🔗 https://nvda.ws/48GoDdi
Download Cosmos: 👉 https://nvda.ws/48Vb3D7
NVIDIA Omniverse: The Future of Industrial Digitalization
I have had several conversations where people have asked me whether NVIDIA Omniverse is just a great renderer. Yes, its rendering capabilities are unmatched, but Omniverse is far more than that. It’s a platform for building, connecting, and orchestrating digital workflows and simulations at scale, acting as a single pane of glass for design, simulation, and operations.
✅ Impact on Industry: Omniverse is transforming how enterprises design, validate, and optimize operations before physical implementation—reducing commissioning risks, accelerating decision cycles, and enabling intelligent automation.
🔑 Key Features:
• Interoperability via USD across CAD, BIM, and simulation tools
• Open & Extensible Architecture for custom workflows
• Real-Time Multiuser Collaboration for global teams
• High Performance & Scalability powered by NVIDIA RTX
• Visual Fidelity for accurate decision-making
• AI & Robotics Integration for automation and synthetic data generation
Read my full article here explaining its key features and how it it emerging as a crucial backbone for the manufacturing and warehouse automation across the industry ⬇️
https://lnkd.in/gQVKYkzz#PhysicalAI#DigitalTwin#NVIDIAOmniverse#Robotics#simulation#PhysicalAIOrchestrator#ManufacturingTransformation#IndustrialAutomation