1 paper
Hongyi Jin, Ruihang Lai, Charlie F. Ruan +5
The recent advances in LLMs bring a strong demand for efficient system support to improve overall serving efficiency. As LLM inference scales towards multiple GPUs and even multipl…