Huawei has unveiled Xinghe AI Cloth 2.0, a major improve to its knowledge middle networking platform initially launched in 2018. The announcement was made at HUAWEI CONNECT 2025’s first Information Middle Innovation Summit, the place greater than 600 world trade leaders, lecturers, and expertise consultants convened to debate the way forward for sustainable, AI-driven infrastructure.
Huawei positions this replace as a basis for creating “always-on” knowledge middle networks designed to totally unlock computing energy whereas supporting the calls for of digital transformation within the AI period.
Throughout his keynote, Arthur Wang, President of Huawei Information Middle Community Area, emphasised that fast advances in synthetic intelligence and the continued evolution of cloud architectures are pushing knowledge middle networks to a crucial turning level. Huawei’s Xinghe AI Cloth 2.0 is constructed round a three-layer structure that features AI Mind, AI Connectivity, and AI Community Components, integrating a variety of recent applied sciences such because the NetMaster community agent, the StarryWing Digital Map automation framework, coaching and inference scheduling engines, and the corporate’s Rock Strong Structure.
Rock Strong Structure
The AI Mind layer brings automation capabilities with drag-and-drop service orchestration, automated deployment, and built-in operations and upkeep throughout heterogeneous networks and a number of safety domains. Huawei says this may considerably enhance operational effectivity for enterprises managing advanced, distributed environments.
The AI Connectivity layer incorporates network-scale load balancing algorithms to lift community throughput to as excessive as 95 p.c and enhance coaching and inference effectivity by greater than 10 p.c. Reliability is additional enhanced by iReliable three-level expertise constructed on the Rock Strong Structure, which the corporate claims delivers as much as ten instances increased reliability than conventional approaches.
On the {hardware} stage, the AI Community Components layer introduces CloudEngine general-purpose computing switches, XH collection clever computing switches, and StarryLink optical modules. These elements are engineered for exact site visitors consciousness, real-time visualization of packet loss and latency, and improved intrinsic safety. Huawei highlights group-based isolation options as one other mechanism to bolster knowledge middle resilience in opposition to rising threats.
With Xinghe AI Cloth 2.0, Huawei is in search of to align knowledge middle community innovation with the accelerating calls for of AI workloads. By combining automation, increased throughput, superior scheduling, and built-in safety, the corporate is concentrating on enterprise and hyperscale operators aiming to maintain efficiency at scale. The discharge displays broader trade momentum towards re-architecting infrastructure for AI-driven functions whereas balancing reliability, effectivity, and safety.
