NVIDIA 用几条命令在 DGX Spark 上运行边缘 AI 模型
原标题：Run AI Models for Edge Application Development with Just a Few Commands
AI 摘要
NVIDIA 发布技术博客，介绍在 64GB 版 DGX Spark 上通过 TensorRT Edge-LLM 的 trtedge CLI，用少量命令即可下载并构建 Nemotron-3.5 Lightning 模型引擎，在终端对话或通过 OpenAI 兼容 API 提供服务，并可用 DSpark 投机解码加速生成。实测开启 DSpark 后生成速度从约 73 tokens/s 提升至 105-127 tokens/s，推理峰值内存 27.9GB。
正文节选
AI is expanding beyond the data center as developers build a new generation of agents to act at the edge: monitoring driver attention and cabin conditions, helping robots understand their surroundings and plan tasks, and coordinating fleets of machines on production lines. These physical AI applications rely on models that perceive, reason, and act within tight memory, power, and latency budgets. But getting a model ready to run can involve substantial setup. NVIDIA DGX Spark, now available in a