#AI #Huawei #Vienna #Austria #Broadband

Transforming Network Value in the AI Era

Recently, at the Broadband Development Congress (BDC) held during Network X 2026 in Vienna, a prominent discussion arose about the strategies needed to adapt existing network structures to better suit the evolving demands of the AI era. Under the spotlight was Bob Chen, President of Huawei's Optical Business Product Line, who emphasized the need for profound changes in how networks are conceptualized and built to leverage the burgeoning capabilities of artificial intelligence.

The Necessity for Transformation

Chen articulated a compelling case for what he describes as the three R's framework: Rebuilding network architecture, Reshaping business models, and Redefining operations and maintenance. This transformative approach aims to unlock new growth potentials that align with the rapid advancements in AI technology. According to Chen, the last three decades have demonstrated that service sector booms—ranging from voice services to internet and video—have always been underpinned by robust network support. In today's reality, AI models are expanding at an unprecedented pace, marking shifts in how networks should be structured and utilized.

As AI models grow larger, with current systems boasting nearly 10 trillion parameters, the demand for flexible and efficient network connectivity has surged, with token consumption increasing by almost 50 times in the past year. As a result, the telecommunications industry faces an urgent necessity to re-evaluate their network value propositions to stay relevant and effective.

Rebuilding Network Architecture

The first vital component in this transformation is Rebuilding the network architecture. Traditionally, networks have been designed around administrative regions, structured hierarchically with national, provincial, and metro segments. However, the rise of Artificial Intelligence Data Centers (AIDCs) has shifted this paradigm dramatically. Chen explained that the volume of network traffic is no longer dictated by population. Instead, it's increasingly influenced by computational power.

Take Zhongwei in China as a case in point; it has transitioned from a region generating merely 0.1% of national traffic to becoming one of China's major national computing hubs. Its intelligent computing needs now account for over 10% of the country’s total bandwidth, indicating a fundamental shift in the network's geographical and structural focus.

Hence, networks need to pivot from being administratively centered to AIDC-centered. This involves creating a layered hierarchy comprising Core Data Centers (DCs), Regional DCs, and Edge DCs, characterized by direct mesh connections that allow for latency cycles ranging from 1 to 20 milliseconds, thus enabling quicker market responsiveness.

Reshaping Business Models

Transitioning to the Reshape phase, Chen highlighted how AI is transforming traditional business models. For instance, businesses involved in industrial design are now leveraging AI rendering to significantly cut production cycles. To cater to such advancements, network operators must evolve their offerings from merely selling bandwidth to delivering integrated solutions that encompass