Let's face it, the standard advice for running local LLMs is to buy the biggest single graphics card you can afford. Local AI ...