--- license: other license_name: nvidia-open-model-license license_link: >- https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-agreement/ --- # Transolver DrivAerML Transolver DrivAerML is a Transformer-based surrogate model for automotive external aerodynamics simulations. It introduces Physics-Attention, which adaptively decomposes the computational domain into learnable physical state slices, enabling efficient attention over physics-aware tokens rather than raw mesh points. The model predicts surface pressure and wall shear stress fields, as well as volumetric velocity and pressure fields on 3D vehicle geometries for computational fluid dynamics (CFD) applications. This model is available for commercial use. ### License/Terms of Use: Use of this model is governed by the [NVIDIA Open Model Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-agreement/). ### Deployment Geography: Global ### Use Case: Computational Fluid Dynamics (CFD) engineers accelerating automotive external aerodynamics with AI. ### Release Date: 05/01/2026 Hugging Face: https://huggingface.co/nvidia/transolver_drivaerml ## Reference(s): [Code](https://github.com/NVIDIA/physicsnemo/tree/main/examples/cfd/external_aerodynamics/transformer_models) [Transolver: A Fast Transformer Solver for PDEs on General Geometries](https://arxiv.org/abs/2402.02366) [DrivAerML: High-Fidelity Computational Fluid Dynamics Dataset for Road-Car External Aerodynamics](https://arxiv.org/abs/2408.11969) ## Model Architecture: **Architecture Type:** Transformer with Physics-Attention over learnable physical state slices. **Network Architecture:** Transolver introduces Physics-Attention, which replaces standard self-attention by first learning soft assignments of mesh points to M physics-aware slice tokens via a linear projection and Softmax. Each of the M tokens is computed as a weighted average of mesh point features. Standard multi-head attention is then applied over the M tokens rather than all N mesh points, achieving O(N) complexity since M ≪ N. Updated token representations are broadcast back to mesh points via the learned slice weights. Each Transolver layer applies LayerNorm, Physics-Attention with residual, LayerNorm, and a feed-forward block with residual. The model comprises 8 such layers. This formulation is grounded theoretically as a learnable integral operator approximation over the PDE domain. **Number of model parameters:** 10M ## Input: **Input Type(s):** - Tensor (3D point cloud coordinates on vehicle surface and volume) **Input Format(s):** PyTorch Tensor
**Input Parameters:** - Point coordinates and geometry features (N, C_g) for surface and volume points, linearly embedded to a common channel dimension C **Other Properties Related to Input:** - Input point cloud represents vehicle surface and volume geometry - Geometry features are concatenated with optional observed physical quantities and linearly projected to the model's working dimension ## Output: **Output Type(s):** Tensor (Surface and volume aerodynamic fields)
**Output Format:** PyTorch Tensor
**Output Parameters:** - Surface: pressure (M_s, 1), wall shear stress (M_s, 3) - Volume: velocity (M_v, 3), pressure (M_v, 1) **Other Properties Related to Output:** - Outputs are normalized using statistics computed from the training dataset - Drag and lift coefficients can be derived via surface integration of pressure and wall shear stress predictions Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA's hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions. ## Software Integration **Runtime Engine(s):** PyTorch
**Supported Hardware Microarchitecture Compatibility:**
* NVIDIA Ampere
* NVIDIA Blackwell
* NVIDIA Hopper
* NVIDIA Turing
**Supported Operating System(s):** * Linux
The integration of foundation and fine-tuned models into AI systems requires additional testing using use-case-specific data to ensure safe and effective deployment. Following the V-model methodology, iterative testing and validation at both unit and system levels are essential to mitigate risks, meet technical and functional requirements, and ensure compliance with safety and ethical standards before deployment. ## Model Version(s): **Model Version:** 1.0.0
# Training, Testing, and Evaluation Datasets: The DrivAerML dataset is used for training and evaluation, which is a publicly available, high-fidelity dataset comprising aerodynamic data for 500 parametrically morphed variants of the DrivAer notchback vehicle. The dataset was generated using hybrid RANS/LES (HRLES), a scale-resolving CFD method, which provides time-averaged quantities for each variant. The available data includes surface pressure, wall shear stress, and flow-field quantities, provided in formats compatible with mesh-based analysis (.vtp for surface data and .vtu for flow-field data). 48 samples (~10%) are used as the test set, with approximately 20% of the test set consisting of out-of-distribution samples based on drag coefficients. These samples represent extreme cases with the lowest and highest drag coefficients in the entire dataset, which remain unseen by the model during training. Models are trained for up to 500 epochs on a single NVIDIA GB200 node using the Muon optimizer. ## Training Dataset: **Data Modality:** - Other: 3D Point Cloud (surface and volume) **Training Data Size:** - 436 files in VTP format (surface meshes) and VTU format (volume flow fields) with corresponding physical quantities **Link:** [DrivAerML Dataset](https://arxiv.org/abs/2408.11969)
*Data Collection Method by dataset:*
* Synthetic CFD Simulation
*Labeling Method by dataset:*
* Automated
**Properties:** The data is a simulation/synthetic dataset generated using hybrid RANS/LES scale-resolving CFD simulations, providing time-averaged surface and volumetric flow fields for different car geometries. Each case contains approximately 150 million volume elements and 10 million surface elements. ## Testing Dataset: **Link:** [DrivAerML Dataset](https://arxiv.org/abs/2408.11969)
*Data Collection Method by dataset:*
* Synthetic CFD Simulation
*Labeling Method by dataset:*
* Synthetic CFD Simulation
**Properties:** Test split from DrivAerML dataset with vehicle geometries held out from training. 48 samples (~10%) are used as the test set, with approximately 20% consisting of out-of-distribution samples based on drag coefficients. ## Evaluation Dataset: **Link:** [DrivAerML Dataset](https://arxiv.org/abs/2408.11969)
*Data Collection Method by dataset:*
* Synthetic CFD Simulation
*Labeling Method by dataset:*
* Automated
**Properties:** Validation split from DrivAerML dataset with vehicle geometries held out from training. The full DrivAerML dataset is split as 90% for training and 10% for validation. ## Inference: **Acceleration Engine:** PyTorch
**Test Hardware:** * H100
* GB200
## Ethical Considerations: NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ subcards: [Bias](https://huggingface.co/nvidia/transolver_drivaerml/blob/main/bias.md), [Explainability](https://huggingface.co/nvidia/transolver_drivaerml/blob/main/explainability.md), [Privacy](https://huggingface.co/nvidia/transolver_drivaerml/blob/main/privacy.md), and [Safety & Security](https://huggingface.co/nvidia/transolver_drivaerml/blob/main/safety.md). Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).