At Network X 2025, Lin Wei, President of the Intelligent Telco Cloud Domain of Huawei Cloud Core Network, delivered a keynote speech titled " From Cloud-Native to AI-Native: Accelerating Telco Intelligence and Innovation." He outlined the emerging challenges that the mobile AI era presents to telecom infrastructure and shared Huawei's latest breakthroughs in core technologies and practices.
As AI continues to advance, it is transforming the industrial landscape faster than ever before. Operators have evolved from mere connectivity providers to enablers of Connectivity+AI services, serving people, vehicles, IoT systems, and agents. In the future, differentiated intelligent experience will define the core competitiveness of telecom services. As this trend continues, Telco Cloud, the backbone of operators' networks, is swiftly advancing from cloud-native to AI-native. At the same time, networks are shifting focus—from pure connectivity to a fusion of connectivity and intelligent computing.
Lin Wei stated that Telco Cloud infrastructure must undergo a comprehensive upgrade—from Cloud-infra to AI-infra. This upgrade hinges on three key changes:
Converging General and Intelligent Computing and Accelerating Containerization
Unified scheduling of AI computing power must be strengthened on Telco Cloud, enabling collaborative allocation of both general and intelligent computing resources. This will ensure the smooth growth of AI services.
Building a Streamlined, High-Performance Network and Computing Power Foundation with a Peer-to-Peer Architecture
The high-speed interconnection bus overcomes traditional CPU-centric limitations by enabling dynamic resource allocation across pooled memory, storage, and NPUs. This architecture ensures optimal data throughput and latency characteristics essential for large-scale AI cluster operations. Such technological advancements facilitate the deployment of robust, streamlined, and elastic infrastructures tailored for the hyper-converged telco cloud infrastructure.
Elevating AI Management and Enabling Capabilities to Support Agile Innovation in AI Services
The AI platform streamlines model management, resource scheduling, and computational acceleration, enabling rapid and seamless deployment of AI solutions. By facilitating the integration of foundational models with domain-specific datasets, they empower organizations to drive AI innovation and achieve tangible business outcomes through accelerated implementation.
Lin Wei said that Huawei's Telco Intelligent Converged Cloud (TICC) is an AI infrastructure solution designed for operators. It helps customers smoothly move to AI-infra in two steps:
Step 1: Follow ETSI standards and introduce bare metal containers with a dual-engine architecture. This approach enables the establishment of a fully integrated cloud platform that seamlessly supports both VMs and containers.
Step 2: As AI services go online, operators should enhance Kubernetes' capabilities, build an AI enablement platform, unify the management of general and intelligent computing, and build a stable, efficient, and easy-to-use AI infrastructure.
In his concluding remarks, Lin Wei said, "Huawei is willing to collaborate with industry partners to jointly shape the future of telco cloud, drive the transformation from cloud-native to AI-native architectures, build next-generation AI-powered telecom infrastructure, and open a new chapter in the mobile AI era."