Skip to content

About

Research: macOS guest with GPU acceleration on Linux/KVM — paravirtual virtio-gpu + Venus (Vulkan) first, VFIO passthrough as fallback; generalised across hardware

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

112 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

macos-gpu-vm-lab

Research notes and tooling for running a macOS guest with GPU acceleration on a Linux/KVM host (AMD Ryzen + NVIDIA GeForce RTX), and for finding out how far that can be generalised to other hardware.

Status: early research. Nothing here is proven to work yet.

Approach

Main track: paravirtual GPU, no passthrough. The guest draws through a virtio-gpu device using Venus (Vulkan over virtio) while the host keeps its GPU. Passthrough is only a fallback/baseline. See docs/PLAN.md for architecture and milestones (M0 to M7).

Rules of this repo

  • No Apple binaries, installers, recovery images or firmware are ever committed. Users fetch macOS themselves.
  • Licence: PolyForm Noncommercial 1.0.0 (personal, research, educational use). This allows reusing code from NullMoth's noncommercial driver as long as its Required Notice is kept (see NOTICE); its LGPL translator stays LGPL.
  • Scripts here are read-only unless their name says otherwise; nothing changes your host config without an explicit step.
  • Apple's EULA restricts running macOS on non-Apple hardware. Personal research use only; you are responsible for your use.

Start

./scripts/check-host.sh

Prints CPU virtualisation flags, IOMMU groups, GPU placement and kernel cmdline. Paste the output in an issue to help map hardware.

See docs/PLAN.md.

About

Research: macOS guest with GPU acceleration on Linux/KVM — paravirtual virtio-gpu + Venus (Vulkan) first, VFIO passthrough as fallback; generalised across hardware

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages