NVIDIA 推出 Personal AI Router 测试版，跨本地设备分发 AI 推理任务
原标题：NVIDIA Personal AI Router Distributes AI Tasks Across Local Compute
AI 摘要
NVIDIA 推出 Personal AI Router（PAIR）测试版，可将局域网内多台电脑的推理能力聚合，自动把 AI 请求分发到各节点，主要面向本地多智能体工作负载。PAIR 通过代理接收请求、识别引擎与模型需求并选择合适节点执行，兼容 Ollama 和 LM Studio，无需改动现有架构。NVIDIA 演示显示，结合 RTX Spark、DGX Spark 和 RTX 5090 可将完成时间缩短约 2 倍，但强调结果取决于工作负载并行度、模型、硬件和网络等因素，并非性能保证。
正文节选
NVIDIA Personal AI Router (PAIR), now available in beta, lets you combine the inference capacity of multiple computers on your local network and automatically distribute AI requests among them. It is primarily designed for local multi-agent AI workloads, where multiple independent model calls can otherwise overwhelm one GPU. NVIDIA says a breadth-first approach to distributing agentic tasks is becoming increasingly common, with a lead agent dispatching subtasks to sub-agents or multiple agents w