I have an old computer that I turned into my ha server with the haos. I have been using it for a while now and am loving it. More recently I got my hands on a few of IP cameras and decided to use motion eye as my NVR. With my new cameras I have 7 in total. this has started to create a large tax on performants. With only 5 install my system is constantly at 99% used. I have some very old GPUs (gt 210 and 710) that I’m not using and though sense most of motion eyes computation is comparing frames for motion detection and encoding video a GPU would greatly improve performants.
Sorry for the long pre-text but I wanted to get that info out first.
So basically, I am writing this to get help and info myself and have this topic for anybody else how wants to do this. I have tried to research this myself but did not find a straight answer.
Q1: Is it possible?
Basically, is it possible to have motion eye run its calculation on a GPU.
I am shore that GPUs can be used by HA since it is commonly used with TensorFlow.
Q2: How?
If what I want is possible how do I install a GPU? What drivers do I use?
Sorry for any spelling or grammatical errors English is my second language.