Credit score: NVIDIA
PRESS RELEASE —NVIDIA in the present day introduced main developments to Common Scene Description, or OpenUSD, that may develop adoption of the common 3D knowledge interchange framework to robotics, industrial design and engineering, and speed up builders’ skills to construct extremely correct digital worlds for the subsequent evolution of AI.
Via new OpenUSD-based generative AI and NVIDIA-accelerated growth frameworks constructed on the NVIDIA Omniverse™ platform, extra industries can now develop purposes for visualizing industrial design and engineering initiatives, and for simulating environments to construct the subsequent wave of bodily AI and robots.
The brand new choices embrace NVIDIA NIM™ microservices for AI fashions that may generate OpenUSD language to reply consumer queries, generate OpenUSD Python code, apply supplies to 3D objects, and perceive 3D area and physics to assist speed up digital twin growth. As well as, new USD connectors to robotics and industrial simulation knowledge codecs and developer instruments let customers stream large, absolutely NVIDIA RTX™ ray-traced datasets to Apple Imaginative and prescient Professional.
“The generative AI increase for heavy industries is right here,” stated Rev Lebaredian, vice chairman of Omniverse and simulation know-how at NVIDIA. “Till just lately, digital worlds have been primarily utilized by artistic industries; now, with the enhancements and accessibility NVIDIA NIM microservices are bringing to OpenUSD, industries of every kind can construct bodily based mostly digital worlds and digital twins to drive innovation whereas making ready for the subsequent wave of AI: robotics.”
Generative AI Involves USD With NVIDIA NIMThe world’s first generative AI fashions for OpenUSD growth, developed by NVIDIA, shall be out there as NVIDIA NIM microservices. The fashions allow builders to include generative AI copilots and brokers into USD workflows, broadening the probabilities in 3D worlds and serving to velocity the adoption of USD throughout a brand new vary of industrial sectors, like manufacturing, automotive and robotics.
The microservices out there in preview are:
USD Code NIM microservice — solutions common information OpenUSD questions and mechanically generates OpenUSD-Python code based mostly on textual content prompts that may then be inputted into an OpenUSD viewing app, akin to usdview from Pixar, or an NVIDIA Omniverse Package-based software to visualise the corresponding 3D knowledge.
USD Search NIM microservice — allows builders to go looking by way of large libraries of OpenUSD, 3D and picture knowledge utilizing pure language or picture inputs.
USD Validate NIM microservice — checks the compatibility of uploaded information in opposition to OpenUSD launch variations and generates a totally RTX-rendered, path-traced picture, powered by NVIDIA Omniverse Cloud APIs, or software programming interfaces.
Newly introduced microservices that shall be out there quickly are:
USD Structure NIM microservice — allows customers to assemble OpenUSD-based scenes from a sequence of textual content prompts based mostly on spatial intelligence.
USD SmartMaterial NIM microservice — predicts and applies a practical materials to a computer-aided design object.
fVDB Mesh Technology NIM microservice — generates an OpenUSD-based mesh, rendered by Omniverse Cloud APIs, from point-cloud knowledge.
fVDB Physics Tremendous-Res NIM microservice — performs AI tremendous decision on a body or sequence of frames to generate an OpenUSD-based, high-resolution physics simulation.
fVDB NeRF-XL NIM microservice — generates large-scale neural radiance fields in OpenUSD utilizing Omniverse Cloud APIs.
Foxconn, a world manufacturing chief with greater than 170 factories worldwide, is already benefiting from NVIDIA’s computing platform, utilizing NIM microservices and Omniverse to create a digital twin of a manufacturing facility below growth.
“Digital twins will assist us speed up the subsequent wave of commercial manufacturing and autonomous machines,” stated Zhe Shi, chief digital officer and head of the Good Manufacturing platform at Foxconn. “NVIDIA Omniverse and the brand new NIM microservices will democratize the flexibility to develop digital twins and assist our groups construct bodily based mostly digital factories quicker than ever.”
WPP, a world chief in advertising and marketing and communications companies firm, is an early adopter of USD Search and USD Code NIM microservices, implementing them in its generative AI-enabled content material creation pipeline, constructed on NVIDIA Omniverse, for patrons akin to The Coca-Cola Firm.
“The fantastic thing about the innovation is how appropriate it’s with the way in which we work, and that it leverages open requirements — accelerating not solely future work, however permitting us to proceed to construct on and lengthen the usefulness of all our earlier investments in requirements like OpenUSD,” stated Stephan Pretorius, chief know-how officer at WPP. “Utilizing NVIDIA NIM microservices with NVIDIA Omniverse has made it attainable for us to launch modern new manufacturing instruments with corporations like The Coca-Cola Firm at unprecedented velocity.”
USD Connectors Carry Generative AI to Extra IndustriesA sequence of recent USD connectors for robotics knowledge codecs and streaming to Apple Imaginative and prescient Professional opens the portals of OpenUSD interoperability and superior authoring to extra industries.
NVIDIA and Siemens, a world chief in industrial automation and software program, are extending their collaboration to facilitate extra industrial workloads utilizing OpenUSD. Siemens will combine OpenUSD pipelines with its Simcenter portfolio of simulation applied sciences to help evidence-based decision-making and collaboration amongst key stakeholders.
This integration allows high-fidelity, real-time, photorealistic visualization of advanced simulation knowledge, offering deeper insights right into a product’s efficiency inside its real-world working surroundings. The work will construct on Siemens’ efforts to include Omniverse into its Teamcenter Product Lifecycle Administration portfolio.
NVIDIA additionally launched a connector from Unified Robotics Description Format to OpenUSD, letting roboticists seamlessly carry their robotic knowledge throughout purposes, together with for design, simulation and reinforcement studying.
To additional advance OpenUSD ecosystem growth, NVIDIA introduced the OpenUSD Alternate software program growth package, enabling builders to construct their very own strong OpenUSD knowledge connectors.
New developer instruments and APIs to stream large-scale OpenUSD scenes from an software constructed on the Omniverse platform to Apple Imaginative and prescient Professional through the NVIDIA Graphics Supply Community are actually out there in early entry.
“OpenUSD is revolutionizing the way in which we create and work together with 3D content material,” stated Steve Might, chief know-how officer of Pixar and chairman of the Alliance for OpenUSD (AOUSD). “Now, with these new companies and APIs for OpenUSD constructed by NVIDIA, we count on to see accelerated progress and adoption of USD, paving the way in which for brand spanking new customers and industries to extra simply have interaction with our ecosystem.”
OpenUSD Ecosystem Momentum