Forecasting LLM Inference Performance via Hardware-Agnostic Analytical Modeling

Open in new window