新闻
DeepSeek-V4.1-Flash on Vera Rubin NVL72: 58x Throughput vs. GB300 NVL72 at 387 tokens/s
vLLM · Inferact and the vLLM Team · 发布于 · 阅读约3分钟
30秒读懂
- 发生了什么
- vLLM achieved 58x higher throughput serving DeepSeek-V4.1-Flash on NVIDIA Vera Rubin NVL72 compared to GB300, reaching 1.7M tokens per second per GPU.
- 为何重要
- Matters for engineers deploying large language models who need maximum inference throughput and latency on latest NVIDIA hardware with specialized tensor cores.
- 注意
- Results are specific to DeepSeek-V4.1-Flash model and SemiAnalysis AgentX benchmark; gains depend on Rubin's split memory domains and FP4 tensor cores, may not generalize to other models.
- agent
- llm
- throughput
- token
- vllm
这条新闻背后的模式
每个模式都讲清楚技术如何运作、何时值得投入，以及在哪里会失效。
The Agent Architect
每周一个模式、一个权衡、一个生产事故案例。为构建智能体系统的人准备的每周简报。
每周一封邮件,一键退订。您的地址仅用于发送简报。