What is edge-desktop-virtualization?

Edge Desktop Virtualization lets you take a single computer with an Intel GPU and split that graphics hardware among multiple virtual desktops running on the same machine. The main benefit is that each virtual machine gets its own dedicated slice of the GPU, so several people can run graphics-heavy applications simultaneously from one physical box, think of it as turning one workstation into several high-performance remote desktops. Under the hood, it uses an Intel technology called SR-IOV, which is a standard for dividing a single physical device into multiple virtual ones. The host operating system manages the full GPU, then carves out isolated virtual "slices" that each virtual machine accesses directly. Because each VM talks to its own slice without going through a software middleman, performance stays high and latency stays low, important for things like video playback, media processing, or any graphics-intensive workload. The project bundles the tools needed to run all of this on Kubernetes-based infrastructure. It includes setup steps for enabling SR-IOV on the host machine, installing Kubernetes along with Kubevirt (a tool for running VMs inside Kubernetes), and an Intel device plugin so the cluster knows how to hand out GPU slices to virtual machines. The README also walks through creating bootable disk images for Windows or Ubuntu guest VMs and deploying them using provided Helm charts. Helm is a package manager for Kubernetes that simplifies deploying applications. The intended users are organizations building edge or cloud desktop infrastructure, for example, a factory floor station serving multiple operator displays from one PC, or a business offering virtual desktops to employees without buying a separate computer for each person. It requires Intel 12th Gen Core processors or newer, with a recommendation of 64GB of RAM and at least 500GB of SSD storage, so this is aimed at fairly capable hardware setups rather than lightweight machines.