An Empirical Evaluation of LLM Inference Efficiency on Data Center GPUs and Edge Accelerators.

Alessandro Costantini, Giovanni Zurlo, Diego Michelotto

Browse the full ICCSA paper archive.