In a computer system, graphic processing unit (GPU) has the responsibility in handling graphical tasks, such as drawing frames to display, animating visual effects or decoding video for playback. When GPU is absent, CPU will need to take on these tasks alongside with its own CPU-bound tasks, which causes strain on the CPU. This is the reason why you experience slow performance and stuttering in the first iteration of VirtualLab@APU: the absent of a GPU.
To help relief CPU workload in a virtual desktop infrastructure (VDI), NVIDIA introduced their virtual GPU technology to help accelerate the virtual workflow of graphical tasks, hence improving user productivity and efficiency. We have deployed their NVIDIA GRID vPC in our current virtual machines, and the difference in performance is staggering.
vGPU in action
You can learn more about Virtual GPU and NVIDIA GRID vPC here.
Watch the following comparison video to see the improvement a virtual GPU can bring to VirtualLab@APU: