CUDA is a parallel computing platform and programming model that enhances computing performance using NVIDIA GPUs. CUDA Deep Learning integrates networking and GPU-accelerated libraries like cuDNN, cuTensor, NCCL, HPC-x, and the CUDA Toolkit.
CUDA Deep Learning
CUDA, developed by NVIDIA, is a parallel computing platform and programming model for GPU. With CUDA, developers can dramatically speed up computing applications by harnessing the power of GPUs. The CUDA Toolkit includes libraries, a compiler, development tools, and the CUDA runtime needed for GPU-accelerated development.
CUDA Deep Learning image extends the CUDA images by adding networking support and additional libraries to accelerate deep learning workloads like cuDNN, cuTensor, NCCL, and HPC-x. These images are provided for use as a base layer upon which to build your own GPU-accelerated application container image.
NEW: The container suite now includes specialized inference-optimized variants (inference-runtime and inference-devel) alongside traditional runtime and devel containers, providing significant size reductions for inference deployment workflows.
Prerequisites
Using the CUDA DL NGC Container requires the host system to have the following installed:
For supported versions, see the Framework Containers Support Matrix and the NVIDIA Container Toolkit Documentation.
No other installation, compilation, or dependency management is required. It is not necessary to install the NVIDIA CUDA Toolkit.
The CUDA Deep Learning NGC Container is also optimized to run on NVIDIA DGX Foundry and NVIDIA DGX SuperPOD managed by NVIDIA Base Command Platform. Please refer to the Base Command Platform User Guide to learn more.
Running Container Using Docker
To run a container, issue the appropriate command as explained in the Running A Container chapter in the NVIDIA Containers For Deep Learning Frameworks User’s Guide and specify the registry, repository, and tags. For more information about using NGC, refer to the NGC Container User Guide. A typical command to launch the container is:
docker run --gpus all -it --rm nvcr.io/nvidia/cuda-dl-base:YY.MM-cuda<xx.y>-devel-ubuntu<YY.MM>
Where:
- YY.MM-cuda<xx.y>-devel-ubuntu<YY.MM> is the container version with "YY.MM" as the release number, "cuda<xx.y>" as CUDA version used in this container and "ubuntu<YY.MM>" as OS this container is built for.
For example:
docker run --gpus all -it --rm nvcr.io/nvidia/cuda-dl-base:24.09-cuda12.6-devel-ubuntu22.04
What Is In This Container?
For the full list of contents, see the CUDA DL Container Release Notes. The NVIDIA CUDA Deep Learning Container is optimized for use with NVIDIA GPUs, and contains the following software for GPU acceleration:
- NVIDIA CUDA
- NVIDIA cuTensor
- NVIDIA cuDNN
- NVIDIA NCCL (optimized for NVLink)
- NVIDIA TensorRT
- NVIDIA HPC-X
Container Variants
Starting with the 26.01 release, the CUDA DL Base container is available in four variants optimized for different use cases:
| Variant | Description | Use Case |
|---|---|---|
| inference-runtime | Minimal inference runtime with stub libraries | Production inference deployment |
| inference-devel | Inference runtime + development headers | Building inference applications |
| runtime | Full CUDA runtime libraries | Production training workloads |
| devel | Full development environment | Development and debugging |
The software stack in this container has been validated for compatibility, and does not require any additional installation or compilation from the end user. This container can help accelerate your deep learning workflow from end to end.
Link to Open Source Code
Security Common Vulnerabilities and Exposures (CVEs)
Please review the Security Scanning tab on NGC to view the latest security scan results. For certain open-source vulnerabilities listed in the scan results, NVIDIA provides a response in the form of a Vulnerability Exploitability eXchange (VEX) document. The VEX information can be reviewed and downloaded from the Security Scanning tab.
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 developer team to ensure this container meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
Please report quality, risk, security vulnerabilities or NVIDIA AI Concerns here.
License
By pulling and using the container, you accept the terms and conditions of this End User License Agreement and Product-Specific Terms.