EAServe: Encode-Aware Disaggregated Serving for Multimodal Large Language Models
EAServe raises multimodal LLM serving goodput by controlling encode, prefill placement, and GPU sharing.
1 stories, newest first · one card per story
EAServe raises multimodal LLM serving goodput by controlling encode, prefill placement, and GPU sharing.
Today 2.6K readers · 2.8K views · 145 reads · all time 206K views