CPU-based stable diffusion

In my previous post about Stable Diffusion and how I could not generate 4k fulldome images with it, I had mentioned running out of RAM as the reason it failed. I thought perhaps if there is a way to run it on CPU instead of GPU, I could use a more recent faster machine - and perhaps use virtual memory and run for a longer time - to generate 4096 x 4096 fulldome images. But trying out rupeshs/fastsdcpu: Fast stable diffusion on CPU - found that even if I do an img2img using a 4K fulldome image as the input, the output was only 1024 x 1024. 

There may be ways to run google colab or something like that, and generate 4K fulldome images - but that is something to be tested later. At first glance, it seems that even Dall-E is limited to 1024x1024, and people have to stitch generated images together to make bigger canvases, as mentioned in this reddit thread. And this thread gives a Stable Diffusion guide which uses upscaling - which has issues as mentioned in my previous post.