Edge Containerization Masterclass - balena Summit 2025 скачать в хорошем качестве

Edge Containerization Masterclass - balena Summit 2025

If you have ever wondered if you can run high-performance applications like Steam inside a Docker container on an edge or embedded device, you can find here the answer. This is an edge containerization masterclass performed during the internal balena Summmit 2025. Many embedded developers believe that applications requiring direct access to a GPU or audio system are not suitable for containerization. In this video, we challenge that assumption and demonstrate how to containerize Steam while maintaining full access to essential hardware components. In this deep-dive tutorial, we explore the process of enabling GPU acceleration, audio playback, interprocess communication and Wayland compositing inside a containerized environment. These techniques are essential for running complex applications on IoT, industrial automation and edge AI devices such as Raspberry Pi, NVIDIA Jetson, Intel x86 and other single board computers powered by balenaOS. We break down key technical challenges and solutions, including: How to provide Docker containers with access to the GPU for hardware-accelerated rendering. The methods to enable ALSA and PipeWire for containerized audio playback. Running Wayland as a separate containerized compositor for graphical applications. Managing persistent storage, device access, and interprocess communication between containers The role of Docker Compose configurations in enabling secure hardware access. By understanding these techniques, engineers and developers working with Docker, Kubernetes, and containerized applications on embedded hardware will gain valuable insights into overcoming the limitations of traditional containerized environments. Why is this important for Edge Computing? As more industrial and IoT applications rely on containerized workloads through Docker containers, understanding how to efficiently manage hardware resources like GPUs, NPUs and audio systems is critical. Whether you're building AI inference pipelines, robotics applications or remote edge devices, the ability to containerize demanding applications is a game-changer. What is balena? balena provides a fleet management platform for deploying and managing Docker containers on edge devices at scale. balenaOS is an optimized Linux distribution designed for running containerized applications on embedded systems, while balenaCloud enables remote fleet management, over-the-air updates, hostOS updates and real-time monitoring for IoT and edge devices. In this project, we leverage balena’s containerized architecture to deploy Steam in an edge environment while maintaining secure, scalable, and reliable system performance. The same techniques can be applied to computer vision, AI inference, industrial IoT, and autonomous systems running on GPUs and NPUs at the edge.