Get Started With the NVIDIA DeepStream SDK

Overview

NVIDIA DeepStream is a powerful SDK that lets you use GPU-accelerated technology to develop end-to-end streaming vision AI pipelines and agents powered by generative AI. The latest release, DeepStream 8, is packed with innovative features to accelerate the development of your next-generation applications.

Release Highlights

Release notes

DeepStream 8.0 Highlights:

• Multi-camera tracking powered by Multiview 3D Tracking (MV3DT) that extends NvTracker to enable distributed, real-time 3D tracking across networks of calibrated cameras with higher accuracy.
• A MV3DT camera calibration tool that calibrates multiple cameras at once against a floor plan, reducing manual effort.
• Inference Builder, which lets you go from YAML config to deployment in minutes with automated pipelines and backend flexibility.
• Support for MaskTracker, our state-of-the-art tracker that generates pixel-level masks in real-time.
• Support for NVIDIA TAO 6 models with Inference Builder or as standalone microservices.
• Support for accelerating the development of generative AI applications

Containers

DeepStream is available in three different flavors of containers:

• Triton: Single container for both x86 and Jetson. (Multi-arch)
• Samples: Single container for both x86 and Jetson.(Multi-arch)
• Development: x86-only container, including Graph Composer
• Arm Development: SBSA container for Arm®-based servers and systems

All Dockerfiles are also available on GitHub. For full details, check the new NGC Collection page.