Comment by dirtyhand

I was considering getting an RTX 5090 to run inference on some LLM models, but now I’m wondering if it’s worth paying an extra $2K for this option instead