Tag
This post discusses budget GPU options (Radeon VII vs two P100s) for LLM inference with an RX 6800, focusing on VRAM vs speed tradeoffs for MoE models.