Custom OS Template for GPU Passthrough

When getting your custom template ready to handle GPU passthrough, be sure to follow the steps laid out here.

Keep in mind, there are two primary scenarios where GPU passthrough comes into play:

• Machine learning tasks (no display required, like OpenCL operations, etc.)
• Image processing workloads (such as OBS Studio, Blender, video editing, or Multiverse presentations, etc.)

While the machine learning scenario demands fewer steps (since the display doesn’t rely on the GPU), this guide walks you through the complete process for both use cases.

Supported GPU

If you’re running DaDesktop on your own hardware or asking for specific cards for a private server, here are a few things to keep in mind.

We primarily test DaDesktop with AMD Radeon RX 6000 series cards (especially the RX 6400) and Nvidia RTX models. Generally, cards from before 2018 aren't supported.

Integrated graphics on Ryzen 7 or 9 processors, as well as Intel iGPUs, should also work, though we don’t carry out automated testing on them.

DD Node side

dadesktop_npnode_deploy/modules/build-os-template includes scripts that tweak the Linux guest to use the GPU as the primary display and toggle between GPU-accelerated setups and CPU-rendered output.

Guest Settings

Please ensure that the "support_qemu" entry in /var/lib/kvm/templatename/sysinfo.json is set to false; this lets you connect to VNC inside the guest, because QEMU's VNC can't display the passthrough GPU screen.

Testing

Method 1: 
vblank_mode=0 glxgears

Method 2:

/apps/dd-guest/check-gpu

Windows guest configuration without a physical monitor (the typical scenario)

We make use of IddSampleDriver to create a display adapter and virtual monitor in the Windows 10 template. If your GPU lacks a physical monitor, just install it.

• Extract the IddSampleDriver package to c:\IddSampleDriver folder

• Install the certificate

• Install the driver

• Navigate to ‘Display settings’ – you’ll see multiple displays. Select ‘Advanced display settings’. Typically, the virtual display we created appears as Display 2 or 3 with the name "Linux FHD". Note the display number (let’s call it N). Next, return to the ‘Select and rearrange displays’ area, choose display N. In the ‘Multiple displays’ section, if the option ‘Show only on N’ is available, enable it and set this display as the main display. Change its resolution .

• If you don’t see a display named "Linux FHD":

Head to ‘Device Manager’ → ‘Display Adapters’ → ‘IddSampleDriver Device’: right-click and enable the device. (You can disable it later if you’re not using the virtual display.)

• If you spot a display named "AMD vDisplay", it’s usually Display 2:

   This display may run slowly; if the option exists, you can "Disconnect this display".

   (Choose Display x, under the “Multiple displays” section, choose the Disconnect this display option from the drop-down menu)

• If you see a display named "Qemu Monitor", it’s usually Display 1:

   You can also "Disconnect this display" if the option is available.