open-gpu-kernel-modules VS MxGPU-Virtualiza

I also love that it can be done with just a few code line changes:

https://github.com/NVIDIA/open-gpu-kernel-modules/commit/1f4...

> I do love AMD because its drivers are open source as opposed to nVidia.

AMD's drivers are not really more open that Nvidia's. Similar to Nvidia's Open GPU Kernel Module's[0], AMD's opensource drivers are mostly a shim that wrap firmware blobs[1] in which the functionality you really care about is contained.

[0] https://github.com/NVIDIA/open-gpu-kernel-modules/discussion...

My understanding is that nowadays most of the heavy lifting is done by magic going on in the firmware, so the actual driver is relatively simple and is open source: https://github.com/NVIDIA/open-gpu-kernel-modules

Check this issue

That's only for a small subset of their more recent GPUs, as you can see here: https://github.com/NVIDIA/open-gpu-kernel-modules#compatible...

Update: it is a driver issue, see https://github.com/NVIDIA/open-gpu-kernel-modules/issues/522

> And I understand NVIDIA is even worse re: GPU virtualization.

Nope, it’s much better on the nvidia side actually. The latest AMD GPU with a publicly accessible GIM driver is the AMD S7150, which was released in 2016. (https://github.com/GPUOpen-LibrariesAndSDKs/MxGPU-Virtualiza...)

> The good news is that I understand this support is actually good on the Intel side

Not anymore. GVT-g is gone on Ice Lake (Intel 10th generation mobile, 11th gen desktop) so that you can no longer do hardware vGPU on newer Intel parts at all.

Sad thing is that what you said used to be true.

Meanwhile NVIDIA GRID needs licensing fees but actually works, with high end GPU options being available.