What starts to become possible with two 3090s that wasn't with just one?

r/LocalLLaMA
Open Source AI

Qwen 3.6 has been working great and has got me wondering. submitted by /u/GotHereLateNameTaken [link] [comments]