Hi KateG ,
Sorry, I did look into this the same day that you had posted it, but I forgot to post back.
It's not that I have an "issue" per se, but just rather that I'm looking to optimize for the future if possible.
To try to pose the question in a more straightforward manner: are there some particular hardware qualities which would be most optimum going forward? For example, if we made sure to get GPU's that support whatever hardware-accelerated encoding that LMI uses?
In a more dense and specific example, the particular host computer I'm using right now has an Nvidia T1000 GPU, and NVENC with that GPU does support various flavors of H.264 and H.265, but it does not support AV1. Other host computers have older GPU's, and NVENC on those only supports H.264.
For what it is worth, I did try switching off the "Display Accelerator" as per the link you had posted. I did then try disconnecting and reconnecting to the host computer, but that didn't seem to change the CPU or GPU load, so I was guessing that perhaps it takes a reboot in order to actually disable the Display Accelerator.
Also for what it is worth, I (anecdotally) notice the increased CPU and GPU usage between multiple different hosts and clients all with different hardware, so I wouldn't think that it is any particular driver. (We also do keep drivers up to date.)
So I will try to find time to test at some point. But basically I am asking a general question about optimizing new hardware purchases for hardware-acceleration of LMI.