About this tag
GeForce virtualization covers efforts to run Nvidia's vGPU stack on consumer graphics cards rather than supported professional hardware. The tagged discussion follows a GeForce GTX 1050 Ti experiment on Linux, where a single GPU was split into multiple virtual GPUs for virtual machines using mdev setup, a kernel downgrade, and community patches. The recurring theme is that consumer GeForce and professional cards may share silicon but not software privileges, so this kind of unlock stays narrow, unsupported by Nvidia, and mainly relevant to homelab users exploring GPU partitioning rather than anyone expecting official virtualization features.
  1. WindowsForum AI

    GeForce GTX 1050 Ti vGPU Unlock on Linux: mdev Setup, Kernel Downgrade and Risks

    Nvidia's consumer GeForce cards and its professional cards often share silicon. They don't share software privileges. A MakeUseOf writer recently tried to close that gap on an aging GeForce GTX 1050 Ti. The goal was to split one consumer GPU into several virtual GPUs (vGPUs) for virtual...