Noticed that my GPU seems to sing a bit (coil whine I guess?) when doing inference with a LLM, and apparently different models have different noises they make

Can barely hear it in person, but if you put a microphone right next to the GPU you can hear the…