Overview of DoMINO-Automotive-Aero NIM#
DoMINO (Decomposable Multi-Scale-Iterative Neural Operator) is a robust, accurate, and scalable AI-based surrogate model that addresses the critical limitations of existing ML approaches for large-scale automotive aerodynamics. Designed for CFD and automotive engineers and researchers, DoMINO significantly reduces automotive aerodynamics simulation times without heavy mesh downsampling or loss of fidelity. DoMINO learns local geometry encodings from point cloud representations and predicts PDE solutions on discrete points using dynamically constructed computational stencils in local regions. By leveraging both short- and long-range geometric features, the DoMINO model accurately predicts solution fields on the vehicle’s surface as well as in the surrounding flow domain—capabilities essential for making design and engineering decisions in industrial applications.
The DoMINO model architecture begins by encoding the geometry (represented as a point cloud) in a global, multi-scale latent space. Next, a subregion or “local geometry representation” is extracted around each point where a solution is desired. Finally, an aggregation network combines the encoded local geometry with relevant stencil-based flow information to produce high-fidelity aerodynamic predictions.
The DoMINO model is part of the NVIDIA PhysicsNeMo, an open-source deep learning framework for building, training, and fine-tuning deep learning models using state-of-the-art Physics-ML methods.
References#
Advantages of NIMs#
NIMs offer a simple and easy-to-deploy route for self-hosted AI applications. Two major advantages that NIMs offer for system administrators and developers are:
Increased productivity — NIMs allow developers to build generative AI applications quickly, in minutes rather than weeks, by providing a standardized way to add AI capabilities to their applications.
Simplified deployment — NIMs provide containers that can be easily deployed on various platforms, including clouds, data centers, or workstations, making it convenient for developers to test, deploy and scale their applications.
The DoMINO-Automotive-Aero NIM provides a fast, accurate model behind a consistent API for automotive external aerodynamic analysis. As part of the broader NVIDIA NIM ecosystem, DoMINO-Automotive-Aero NIM can be used in conjunction with other NIMs to build pipelines that allow acceleration of vehicle design optimization.
You are responsible for ensuring that your use of NVIDIA AI Foundation Models complies with all applicable laws.