Kasm Workspaces - GPU Passthrough

This video demonstrates how to configure GPU PCI passthrough on Harvester and enable Nvidia GPU support for Kasm Workspaces. Learn how to install Nvidia drivers, configure the Nvidia container runtime, and verify GPU visibility within Kasm. It also showcases AI and ML-ready workspaces from the Kasm AI Registry, including Easy Diffusion, CUDA-enabled PyTorch, and Blender with GPU acceleration. Kasm GPU documentation: https://kasmweb.com/docs/latest/how_to/gpu.html CHAPTERS 00:00 Intro 00:36 Setup GPU PCI Passthrough on Harvester 03:49 Installing Kasm 04:27 Install NVIDIA Drivers and NVIDIA Container Toolkit 05:56 Installing Workspaces 06:06 Installing Kasm AI Registry 06:41 Enabling GPU for Workspaces 07:20 Verify GPU compatibility on Kasm 07:55 Live Demo: Ubuntu Noble with GPU support 08:37 Live Demo: Easy Diffusion Workspace with GPU support 09:13 Live Demo: CUDA enabled PyTorch with GPU support 09:45 Live Demo: Blender Workspace with GPU support Helpful Links: FAQ: https://www.kasmweb.com/docs/latest/FAQ.html Release Notes: https://www.kasmweb.com/docs/latest/release_notes.html How To Guides: https://www.kasmweb.com/docs/latest/how_to.html Sizing and Deployment Guide: https://www.kasmweb.com/docs/latest/how_to/sizing_operations.html User Guide: https://www.kasmweb.com/docs/latest/user_guide.html Administration Guide: https://www.kasmweb.com/docs/latest/admin_guide.html Installation Guide: https://www.kasmweb.com/docs/latest/install.html Admin Account Recovery: https://www.kasmweb.com/docs/latest/how_to/admin_account_recovery.html Kasm Github: https://github.com/kasmtech Kasm Docker Hub Repositories: https://hub.docker.com/u/kasmweb