NVIDIA Modulus
NVIDIA Modulus is an open-source framework for building, training, and fine-tuning Physics-ML models with a simple Python interface.
Using Modulus, engineers can build high-fidelity AI surrogate models that blend the causality of physics described by governing partial differential equations (PDEs) with simulation data from CAE solvers or observed data. Such AI models can predict with near-real-time latency and for a parameterized design space.
Using Modulus, you can bolster your engineering simulations with AI. You can build models for enterprise scale digital twin applications across multiple physics domains, from CFD and Structural to Electromagnetics.
Benefits
AI Toolkit for Physics
Configure, build, and train AI models for physical systems quickly with simple Python APIs.
The framework is generalizable to different domains—from engineering simulations to life sciences and from forward simulations to inverse/data assimilation problems.
Customize Models
Download and customize pretrained state-of-the-art models from the NVIDIA NGC? catalog. Build on the reference applications and extend it to your use case.
Near-Real-Time Inference
Deploy AI surrogate models as digital twins of your physical systems to simulate in near real time.
Scale With NVIDIA AI
Leverage NVIDIA AI to scale training performance from single-GPU to multi-node implementations.
See Modulus in Action
Accelerating Extreme Weather Prediction with FourCastNet
Siemens Energy HRSG Digital Twin Simulation Using NVIDIA Modulus and Omniverse
Maximizing Wind Energy Production Using Wake Optimization
Key Features
New Model Architectures
Modulus offers a variety of approaches for training physics-based models, from purely physics-driven models like PINNs to physics-based, data-driven architectures such as neural operators.
Modulus includes curated Physics-ML model architectures, Fourier feature networks, or Fourier neural operators trained on NVIDIA DGX across open-source, free datasets found in the documentation.
Training State-of-the-Art Physics-ML Models
Modulus provides an end-to-end pipeline for training Physics-ML models—from ingesting geometry to adding PDEs and scaling the training to multi-node GPUs. Modulus also includes training recipes in the form of reference applications.
Explicit Parameterization
Modulus provides explicit parameter specifications for training the surrogate model with a range of values to learn for the design space and for inferring multiple scenarios simultaneously.
Omniverse Integration
Modulus is now integrated with NVIDIA Omniverse? via the Modulus extension that can be used to visualize the outputs of a Modulus-trained model. The extension enables you to import the output results into a visualization pipeline for common output scenarios, such as streamlines and iso-surfaces. It also provides an interface that enables interactive exploration of design variables and parameters for inferring new system behavior and visualizing it in near real time.
Ways to Get Started With NVIDIA Modulus
Download Containers and Models for Development
Develop Physics-ML models using Modulus container and pretrained models, available for free on NVIDIA NGC.
Download Now
Enterprise-Scale Workflows
Get free access to NVIDIA cloud workflows for Modulus and experience the ease of scaling to enterprise workloads.
Try on NVIDIA LaunchPad
Self-Paced Online Course
Take a hands-on introductory course from the NVIDIA Deep Learning Institute (DLI) to explore physics-informed machine learning with Modulus.
Access Course
What Others Are Saying
Modulus Featured Content
Download the NVIDIA Modulus Framework Today