Transforming Network Value for New Growth in the Age of AI

In a significant event, the World Broadband Association (WBBA) hosted the Broadband Development Congress (BDC) during the Network X 2026 conference in Vienna. Bob Chen, President of Huawei's Optical Business Product Line, delivered an insightful keynote titled "Transforming Network Value for New Growth in the Age of AI". This speech underscored the imperative shift in network building paradigms to accommodate the rapid advancements in AI technology, which is characterized by three essential pillars: Rebuilding, Reshaping, and Redefining.

Bob Chen reflected on the service booms of the past three decades—from voice to internet and now to video—each dependent on robust network support. The current AI landscape is evolving exponentially, with large AI models now encompassing nearly 10 trillion parameters and token consumption skyrocketing by almost fifty times within a year. The introduction of AI has brought about dramatic shifts in networking, including the nature, frequency, and quality of connections, prompting a critical reassessment of how to transform network value effectively.

Key Strategies for Transformation

1. Rebuilding Network Architecture

The cornerstone of transforming network value lies in revising the very conceptualization of network infrastructure. Historically, network constructs were organized into administrative regions, segmented into national backbone networks, provincial backbones, and metropolitan networks. However, with the surge in AI usage, there has been a tenfold increase in Artificial Intelligence Data Centers (AIDC), demanding a shift in focus from population-based to compute-based metrics. An example of this shift can be seen in Zhongwei, China, which has transitioned from accounting for just 0.1% of the country’s total traffic to becoming one of China's eight national computing nodes, now representing more than 10% of the total bandwidth.

As AI continues to flourish, the network architecture must pivot from a region-based structure to one centered around AIDC. It is crucial that networks are organized in hierarchical layers surrounding core data centers (DCs), regional DCs, and edge DCs, facilitating swift, deterministic latency cycles ranging from 1 to 20 ms for seamless connectivity.

2. Reshaping Business Models

With AI's influence permeating the enterprise sector, business models are undergoing transformative modifications. For instance, businesses utilizing AI in industrial design can leverage AI-rendering techniques to drastically shorten production cycles, necessitating latency below two milliseconds. This evolution invites operators to extend their offerings from traditional bandwidth to a multi-faceted approach combining bandwidth, latency, and reliability alongside integrated network-computing services.

In the realm of home broadband, the focus is shifting from simply providing broadband access to delivering comprehensive network coverage within residential environments, ultimately moving towards the development of smart home ecosystems. This entails offering high-quality Wi-Fi networks throughout homes and enabling rich AI-driven applications, transforming user interactions with broadband connections via real-time vocal commands.

3. Redefining Operations and Maintenance

In an age heavily influenced by AI, there is an urgent need to incorporate comprehensive intelligence in the planning, construction, maintenance, optimization, and operational phases of networking. For instance, a significant percentage of home network issues stem from Wi-Fi-related problems. Currently, AI-enabled Wi-Fi tuning systems can identify interference sources automatically, adjusting broadcasting power and other parameters in real-time, improving Wi-Fi performance by about 20% in congested environments.

Additionally, devices previously operated at full capacity without regard for actual utilization can cause excessive energy consumption. Advanced traffic intelligence now permits the dynamic regulation of energy usage for network equipment, achieving reductions of over 20%. Moreover, intelligent self-healing capabilities are being developed to address broadband connectivity issues across three tiers: automatic resolution for wireless interface problems, remote management of configuration issues, and precise localization of physical connection issues facilitated by AI agents.

Conclusion

As we stand on the brink of the AI era, Bob Chen's compelling insights underscore the necessity for a paradigm shift towards rebuilding network architectures, reshaping business models, and redefining operational protocols. Through these fundamental reforms encapsulated in the three R's, the broadband landscape can adapt to foster significant growth in the AI-driven age.