What Is DeepStream?

DeepStream is NVIDIA's GPU-accelerated toolkit for building real-time video, image, and sensor analytics pipelines. This page hosts the DeepStream containers for x86, ARM SBSA, DGX Spark, and Jetson.

For the full picture across the release, see the DeepStream SDK Collection.

Available containers

Read the DeepStream 9.1 Release Notes for the full list of features in this release.

Note. DeepStream containers derived from previous releases (before DeepStream 6.1) need to update their CUDA GPG key to perform software updates. See the CUDA Linux repository key rotation notice for details.

Getting Started

Choose your platform below. The pull command is the same for every platform; only the run flags differ.

Prerequisites

Pull the container

1. From the top-right of this page, select the pull-down Get Container and copy the URL for the default container. Or click View all tags to select a different container.

2. Pull the container. Ensure the pull completes before continuing:

docker pull nvcr.io/nvidia/deepstream:9.1-triton-multiarch

Run the container

First, allow external applications to connect to the host's X display:

xhost +

Then run the container using the block matching your platform.

On x86:

docker run --gpus all -it --rm --network=host --privileged -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -w /opt/nvidia/deepstream/deepstream-9.1 nvcr.io/nvidia/deepstream:9.1-triton-multiarch

On Jetson:

docker run -it --rm --runtime=nvidia --network=host -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix -w /opt/nvidia/deepstream/deepstream-9.1 nvcr.io/nvidia/deepstream:9.1-triton-multiarch

Additional arguments for cameras on Jetson:

• CSI camera: -v /tmp/argus_socket:/tmp/argus_socket
• USB camera: --device /dev/video<N>

On ARM SBSA and DGX Spark:

docker run -it --rm --runtime=nvidia --network=host -e NVIDIA_DRIVER_CAPABILITIES=compute,utility,video,graphics --gpus all --privileged -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix -v /etc/X11:/etc/X11 nvcr.io/nvidia/deepstream:9.1-triton-sbsa-dgx-spark

Docker command line options explained:

Notes:

General (applies to all platforms):

• Refer to /opt/nvidia/deepstream/deepstream-9.1/README inside the container for details on deepstream-app usage.

• Using the Triton docker as a base image. For creating a base image using the Triton docker as a baseline, one approach is to use an entry point with a combined script so end users can run a specific script for their application. Add the following to your Dockerfile:

ENTRYPOINT ["/bin/sh", "-c", "/opt/nvidia/deepstream/deepstream-9.1/entrypoint.sh && <custom command>"]

• For Triton samples, while running /opt/nvidia/deepstream/deepstream-9.1/samples/prepare_classification_test_video.sh, the FFMPEG package along with additional dependent libs need to be installed using the command below. For additional information refer to section 1.2 (codecs: DIFFERENCES SINCE DEEPSTREAM 6.1 AND ABOVE) and section 1.3 for BREAKING CHANGES in the Release Notes.

apt-get install --reinstall libmp3lame0 libxvidcore4 ffmpeg

On Jetson:

• These containers use the NVIDIA Container Runtime for Jetson. The runtime exposes device-specific BSP resources to the container so DeepStream can access the hardware it needs at runtime.

• JetPack 6.0 and newer. The NVIDIA Container Runtime no longer mounts user-level libraries (CUDA, cuDNN, TensorRT) from the host. These libraries are installed inside the DeepStream container instead.

• Building custom Jetson containers. See the Docker Containers section of the DeepStream Developer Guide.

Technical Support

DeepStream Support is available via:

Known Limitations

DeepStream 9.1 Triton containers (x86 and SBSA) have the following known CVEs:

License

The following licenses apply to the DeepStream SDK assets:

NOTE: By pulling, downloading, or using the DeepStream SDK, you accept the terms and conditions of the EULA licenses listed above.
This project will download and install additional third-party open source software projects. Review the license terms of these open source projects before use.

Please note that all container images come with the following packages installed:

• librdkafka

• hiredis

• cmake

• autoconf ( license and license exception )

• libtool

• libglvnd-dev

• libgl1-mesa-dev,libegl1-mesa-dev,libgles2-mesa-dev

The software listed below is provided under the terms of GPLv3.

To obtain source code for software provided under licenses that require redistribution of source code, including the GNU General Public License (GPL) and GNU Lesser General Public License (LGPL), available under DeepStream Resources 9.1_3rd_party_oss. This offer is valid for a period of three (3) years from the date of the distribution of this product by NVIDIA CORPORATION.

Get Help

Enterprise Support

Get access to knowledge base articles and support cases or submit a ticket.

Ethical AI

NVIDIA’s platforms and application frameworks enable developers to build a wide array of AI applications. Consider potential algorithmic bias when choosing or creating the models being deployed. Work with the model’s developer to ensure that it meets the requirements for the relevant industry and use case; that the necessary instruction and documentation are provided to understand error rates, confidence intervals, and results; and that the model is being used under the conditions and in the manner intended.