NVIDIA unveiled the NVIDIA OmniverseTM Blueprint, which allows industry software developers to assist their computer-aided engineering (CAE) clients in the industrial, energy, automotive, aerospace, and other sectors in creating interactive digital twins in real-time.
Software developers like Altair, Ansys, Cadence, and Siemens can use the NVIDIA Omniverse Blueprint for real-time computer-aided engineering digital twins to assist their clients in reducing development costs and energy consumption while accelerating time to market.
Also Read: NVIDIA and SoftBank Corp. Partners to Quicken Japan’s Transition to a Global AI Powerhouse
To accomplish 1,200x quicker simulations and real-time visualization, the blueprint is a standard approach incorporating physics-AI frameworks, NVIDIA acceleration libraries, and interactive, physically based rendering.
The creator and CEO of NVIDIA, Jensen Huang, stated in the official blog, “We built Omniverse so that everything can have a digital twin. Omniverse Blueprints are reference pipelines that link AI technologies with NVIDIA Omniverse, allowing top CAE software developers to create innovative digital twin workflows that will revolutionize industrial digitalization for the biggest industries in the world, from design and manufacturing to operations.”
Computational fluid dynamics (CFD) simulations, a crucial first step in virtually exploring, testing, and improving the designs of automobiles, aircraft, ships, and numerous other goods, are among the earliest uses of the blueprint. It can take weeks or even months to finish traditional engineering operations, which include physics simulation, visualization, and design optimization.
A virtual wind tunnel that enables users to simulate and visualize fluid dynamics at real-time, interactive speeds—even when altering the vehicle model inside the tunnel—is being demonstrated by NVIDIA and Luminary Cloud at SC24, marking an industry first.
Also Read: Nvidia Eyes Partnership with India for Chip Development, Targeting Expanding AI Market
Unifying NVIDIA’s Three Technology Foundations for Developers
Real-time physics solver performance and real-time visualization of large-scale datasets are two essential skills needed to build a real-time physics digital twin.
In order to accomplish these, the Omniverse Blueprint combines the NVIDIA CUDA-XTM libraries to speed up the solvers, the NVIDIA Modulus physics-AI framework to train and implement models to create flow fields, and the NVIDIA Omniverse application programming interfaces for real-time RTX-enabled visualization and 3D data interoperability.
The blueprint can be fully or partially integrated into the developers’ current tools.
Ecosystem Advances Simulations with NVIDIA Blueprint
To facilitate rapid CFD simulation, Ansys was the first to implement the Omniverse Blueprint in its Ansys Fluent Fluid Simulation program.
At the Texas Advanced Computing Center, Ansys used 320 NVIDIA GH200 Grace Hopper Superchips to run Fluent. After just over six hours, a 2.5-billion-cell car simulation that would have taken over a month to execute on 2,048 x86 CPU cores was finished. This greatly increased the viability of overnight high-fidelity CFD assessments and set a new industry standard.
The model is also being used by Luminary Cloud. Using training data from its GPU-accelerated CFD solver, the company’s new simulation AI model—which is based on NVIDIA Modulus—learned the connections between airflow fields and vehicle shape.
Also Read: Denmark’s First AI Supercomputer Gefion Launched by Jensen Huang of Nvidia & Danish King
Through the use of Omniverse APIs, real-time aerodynamic flow simulation is made possible by the model, which does simulations orders of magnitude quicker than the solver itself.
Siemens, SimScale, Altair, Beyond Math, Cadence, Hexagon, Neural Concept, and Trane Technologies are also investigating the possibility of incorporating the Omniverse Blueprint into their systems.
All of the top cloud computing systems, such as Amazon Web Services, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure, are compatible with the Omniverse Blueprint. Additionally, NVIDIA DGXTM Cloud offers it.
The NVIDIA Omniverse Blueprint is being used by Rescale, a cloud-based platform that helps businesses speed up scientific and engineering discoveries, to make it possible for businesses to train and implement unique AI models with a few clicks.
The Rescale platform can be used with any cloud service provider and automates the entire application-to-hardware stack. Businesses can use any simulation solver to create training data, construct, train, and implement AI models, run inference predictions, and display and optimize models.
Accessibility
Businesses can register for early access to the NVIDIA Omniverse Blueprint for real-time digital twins in computer-aided engineering.
Also Read: NVIDIA NVLM 1.0: Know All About Open-Source Multimodal LLM