NVIDIA发布Groq 3 LPX推理芯片，Vera Rubin系统全面投产
NVIDIA 推出新一代推理芯片，专为 agentic AI 优化，长上下文生成速度达竞品 4 倍，推理正成为 AI 新战场。
NVIDIA 宣布其机架级系统 Groq 3 LPX 已全面投产，专为 agentic AI（自主决策型 AI，能独立执行多步任务）设计。在人工分析基准测试中，处理 10 万 token 长上下文时每秒可生成 3400 个输出 token，比最接近的竞品快 4 倍。SpaceXAI 和 CoreWeave 等伙伴已开始采用相关平台。
正文摘录
The next era of AI inference won’t be defined by a single breakthrough chip, network or system. It’ll be defined by how every layer of the AI factory works together. That’s why NVIDIA is extending Vera Rubin NVL72 with fast token generation for agentic systems. [Announced today](https://nvidianews.nvidia.com/news/nvidia-groq-3-lpx-now-in-full-production-with-world-class-speed-for-agentic-ai) , the NVIDIA Vera Rubin rack-scale system NVIDIA Groq 3 LPX is in full production. In an Artificial Analysis benchmark running Gemma 4 31B, an open source agentic model, it delivered 3,400 output tokens per second for 100,000-token long-context use cases critical to agentic systems, 4x faster than the nearest alternative platform. Industry partners worldwide are adopting Vera Rubin platform solutions. …