MirkoCovizzi
commented
5 years ago I'm very well aware, in fact it is in the notes of the features working/not working under the dedicated GPU section. A user on tonymacx86 was able to get it working on his Thinkpad P71, however that laptop has a different configuration than the P50, P51 and also P72.
https://www.tonymacx86.com/threads/guide-getting-nvidia-cards-to-work-on-thinkpad-p71.274870/
This is the closest we can get. The dedicated GPU and iGPU are definitely muxed since both can connect directly to the device display (not through optimus but a real direct hardware connection), however the way the operating systems understand how the dedicated GPU is setup is probably through BIOS EEPROM. Mac OS gets this info directly from the GPU EEPROM usually, not from the BIOS. So, the information is there, we just need a way, if possible, to supply the information (information we extract from BIOS) to Mac OS on boot letting it think that we have a separated GPU connected to the motherboard.
I'm currently investigating a probably better alternative for having Mac OS on this laptop, which is through KVM, i.e. through virtualization. In theory we can use dedicated GPU pass-through for graphic acceleration. This way we would have a virtual Mac OS easier to maintain with just a performance drop of 5% on both CPU and GPU.
nicosrs
As you know it is possible to use the dedicated Nvidia GPU for the external ports. But when disabling "Hybrid Graphics" in the BIOS, could the internal screen work using the Nvidia GPU as well? Probably not, but according to this post it looks like the integrated GPU is not active at all anymore (Zbook 15, similar hardware setup): https://www.tonymacx86.com/threads/guide-hp-zbook-15-g2-nvidia-sierra-clover-journey-in-progress.206031/post-1402257