At the Broadband Development Congress during Network X 2026 in Vienna, Huawei Optical Business Product Line President Bob Chen outlined a strategy for adapting network infrastructure to meet the demands of artificial intelligence. Chen argued that operators must alter their network deployment philosophy to support surging AI workloads.
The Three R's of AI Networking
According to Huawei, adapting networks for the AI era requires applying three principles: rebuilding network architectures, reshaping business models, and redefining network operations and maintenance. Chen stated that large AI models now feature nearly 10 trillion parameters, while token consumption increased nearly 50-fold over the past year.
To support this growth, Huawei recommends shifting network architecture from regional administration to AI Data Center (AIDC) centric designs. This transition includes organizing networks into hierarchical layers around core, regional, and edge data centers to achieve deterministic latencies of 1, 5, and 20 milliseconds.
Business Models and Maintenance
In the enterprise segment, Huawei suggested that operators evolve from selling simple bandwidth to offering multidimensional services combining bandwidth, latency, and reliability alongside integrated computing and networking. For home broadband, the company proposed moving toward whole-home Wi-Fi and smart home experiences powered by AI agents.
Huawei also highlighted that AI-based WLAN optimization can automatically identify interference sources and adjust parameters in real time. According to the company, this capability increases Wi-Fi speeds by about 20 percent in disruptive environments, while intelligent traffic analysis reduces overall device energy consumption by over 20 percent.
Key facts and where they come from
- Bob Chen stated that large AI models now have nearly 10 trillion parameters.
Große KI-Modelle verfügen mittlerweile über fast 10 Billionen Parameter
- Token consumption increased by nearly 50 times over the past year.
der Token-Verbrauch ist im vergangenen Jahr um fast das 50-Fache gestiegen.
- AI-based WLAN optimization increases Wi-Fi speeds by about 20 percent in interference environments.
wodurch die WLAN-Geschwindigkeiten in Umgebungen mit Störungen um etwa 20 % gesteigert werden.
- Intelligent traffic analysis reduces overall device energy consumption by over 20 percent.
wodurch der Gesamtenergieverbrauch der Geräte um über 20 % gesenkt wird.
Read the original from Company press releases (PR Newswire) →