Hello Everyone!
New here to LocalLLM.
Looking to setup my first Local LLM. I currently have a 5090 (32 GB VRAM) in my main system, I also have another spare 5080 (16GB VRAM) in a 2nd pc that I can source.
I only have 32GB DDR RAM tho. Running on i9-12900k
From some research it looks like I should start with OpenCode + vLlama(?) + Qwen 3.6 27GB or 35GB MOE model.
Questions:
Should I just run off 1 5090 and be done with with?
What extra performance can I gain adding a 5080? Should I ever bother?
being 2 different GPU, should I even other have 2 GPU runs?
If anyone can help me with some optimed setup/parameters/config for both setup I'd be forever grateful.
I'll probably will have more questions as I time goes on, but just hopeful to get these answered for now.