LLM Inference

or “How to serve 1 Billion Tokens per day in a 45 kW rack”