vLLM · 官方博客

呵

PD Serving of Qwen3.8-2.4T

PD Serving of Qwen3.8-2.4T

二〇二六年九月二十一日 · 英文原文
摘要

How vLLM reaches 5K throughput and 180 interactivity on Qwen3.8-2.4T with GB300 NVL72 PD serving and how to reproduce results yourself.

How vLLM reaches 5K throughput and 180 interactivity on Qwen3.8-2.4T with GB300 NVL72 PD serving and how to reproduce results yourself.

译自 vLLM · 官方博客 · 录于 二〇二六年九月二十一日