Core Insights - The article discusses Huawei's launch of the AI UBB solution aimed at transforming ultra-broadband networks to stimulate new business growth through AI technology [1][7] - The solution encompasses four main components: AI FAN, AI OTN, AI WAN, and ADN, providing end-to-end coverage from home networks to carrier networks [1][3] Group 1: AI UBB Solution Components - AI FAN upgrades home gateways to a unified AI entry point, enabling seamless interaction through various modalities such as voice and video [3][4] - AI OTN enhances carrier networks by providing high-quality transport capabilities through intelligent analysis of business flow characteristics, ensuring low latency and high bandwidth [4][5] - AI WAN focuses on the precise identification of AI business types, including encrypted streams, and offers differentiated service packages to ensure optimal user experience [5][6] Group 2: Technological Advancements - The evolution of general artificial intelligence is accelerating, with significant advancements in large models and multi-modal capabilities, leading to a dramatic reduction in inference costs by approximately 100 times over the past two years [3] - The shift from traditional networks to intelligent networks is facilitated by the integration of edge computing and AI technologies, enhancing the overall performance and security of broadband services [6][7] Group 3: Strategic Vision - Huawei aims to collaborate with the industry to establish unified standards for AI UBB, fostering a mature application ecosystem and exploring new business models [7] - The ultimate goal is to achieve a high level of automation in network management, significantly reducing the need for manual intervention in network operations [7]

[中国，上海，2025年6月19日] 在2025 MWC 上海期间，全球云网宽带产业协会WBBA举办了上海BDC峰会。在峰会上，华为常务董事汪涛发表了"AI全面重 构超宽带（UBB）网络，激发商业新增长"主题演讲，并推出AI UBB解决方案，包括AI FAN、AI OTN、AI WAN和ADN四大部分，端到端覆盖家庭网络、 承载网络和UBB智能管控系统。AI UBB解决方案通过端到端设备内生智能和全面能力提升，为AI应用提供极致体验，实现高品质入算和高效算力互联，加 速网络迈向高阶自智，激发商业新增长。 华为常务董事汪涛发表主题演讲 当前，通用人工智能时代正在加速到来。大模型正在从LLM走向多模态，Agent的能力和MCP（模型上下文协议）、A2A（智能体到智能体协议）等使能技 术正在快速发展，推动AI行业应用成熟度快速提升。MOE（混合专家模型）、模型蒸馏、高性能算力的发展使得推理成本快速下降，过去两年LLM的推理 成本下降了约100倍。高性能的小模型和端云协同技术支撑AI正在加速进入智能手机、眼镜、PC等各种终端。 AI OTN：内生智能和光电协同能力提升，为AI应用和算力协同提供高品质运力 更低的时延是 ...