1/3
3 Years of Service
I've used ComfyUI on an AMD card before, so it was rendering on CPU. I don't think that ComfyUI is your problem, if you already got it running just try a different checkpoint or try someone else's workflow.So who is using a 6 vram cpu or less for local AI gen? What do you use to get decent images? I tried Comfy and was able to do some terrible img2vid but none of my text2img come out at all.
But it can be sloooooow