Linux Docker Setup: How to Assign a Spare GPU to Containers

Learn how to run NVIDIA Docker containers on a secondary GPU (GPU2) in a dual GPU Linux system.
Keep GPU0 free for desktop or gaming while GPU2 handles Docker, CUDA, and AI workloads.

This video shows a step-by-step NVIDIA Docker setup on a secondary GPU (GPU2) using a dual GPU Linux system.
If you have multiple NVIDIA GPUs, this guide helps you:
- Isolate Docker workloads to GPU2
- Keep GPU0 free for desktop, gaming, or display
- Run CUDA, AI, and ML containers without conflicts

What you’ll learn:
-Install NVIDIA Container Toolkit on Linux
-Verify GPU access inside Docker containers
-Run Docker containers only on a specific GPU (GPU2)
-Enter the container and validate GPU usage using nvidia-smi

Setup Ideal for:
- Dual GPU Linux users
- AI / ML developers
- Docker + CUDA workflows

Running Docker on a secondary GPU prevents desktop lag, avoids Xorg conflicts, and gives you clean GPU isolation for compute workloads.

0:00 Intro
0:29 Dual GPU Configurations
1:37 Install and configure Docker
2:53 Install and configure Nvidia Container Toolkit
3:50 Run Nvidia Container image with GPU-2
7:02 Outro
If this channel helps you, you can support it here ☕
https://buymeacoffee.com/TheIndianPengu

#NvidiaDocker #DockerGPU #DualGPU #Linux #CUDA #AI

Видео Linux Docker Setup: How to Assign a Spare GPU to Containers канала The Indian Pengu