How NVIDIA Groq 3 LPX Unlocks Ultrafast Interactivity at Long Context on NVIDIA Vera Rubin

（翻译）NVIDIA Groq 3 LPX 如何在 NVIDIA Vera Rubin 上实现长上下文的超快交互

资讯摘要

NVIDIA Groq 3 LPX is the interactive AI inference accelerator for the NVIDIA Vera Rubin platform. At the core of the platform is NVIDIA Vera Rubin NVL72…

AI 摘要

NVIDIA技术博客介绍了Groq 3 LPX推理加速器，它与Vera Rubin NVL72平台结合，通过编译器调度、点对点C2C链路和细粒度计算通信重叠，降低首比特延迟，使模型能以每秒3000+ token的速度在100K上下文下运行。第三方Artificial Analysis基准显示，Gemma 4 31B模型在Groq 3 LPX上达到3,431 tokens/s输出，支持长上下文多轮智能体会话。