GVT G vs GVT D

What is the difference between GVT-G and GVT-D?

Both are a technology implemented on intel GPUs that allows for the enhanced graphics performance on the guest operating system in a virtualization scenario. They are different in hwo this is exactly done.

GVT-G

GVT-G creates a virtual GPU device that has a limited subset of the performance, memory, and resolution capabilities of the physical iGPU. It still allows for the Host OS to use the GPU or for multiple guests to use the GPU simultaneously(to a limit, as each one only has so many vgpus, see table below). This ROM creates the necessary configuration and framebuffer setups to allow these to generate an output that QEMU and redirect to the appropriate display server. Table to Come

GVT-D

GVT-D is essentially traditional GPU passthrough. The entire iGPU is given to the guest OS, including the physical outputs. This ROM initializes all the components of the chip and prepares it for output. It also now allows for the use of both regular and embedded DisplayPort devices including laptop screens(Even having the possibility brightness control!). This setup means that the guest has exclusive access to the gpu resources, and neither hosts nor other guests can utilize it.