ZTE is closely investing in AI-native community structure and high-efficiency computing
In sum – what to know:
AI-native networks – ZTE is redesigning telecom infrastructure with embedded intelligence, modular knowledge facilities and AI-driven vitality programs to fulfill escalating AI compute calls for.
Smarter fashions, higher knowledge – With superior coaching strategies and instruments, ZTE is pushing the boundaries of mannequin accuracy and real-world efficiency.
Edge-to-core deployment – ZTE is integrating AI throughout client, house and business use instances, delivering over 100 utilized options throughout 18 verticals below its “AI for All” technique.
SHANGHAI — ZTE’s Chief Improvement Officer Cui Li outlined the seller’s roadmap for constructing clever infrastructure and accelerating synthetic intelligence (AI) adoption throughout industries throughout a keynote session at MWC Shanghai 2025.
Throughout her speech, Cui highlighted the rising affect of enormous AI fashions and the essential function of foundational infrastructure. “Irrespective of how AI know-how evolves sooner or later, the main target will stay on environment friendly infrastructure, optimized algorithms and sensible functions,” she mentioned.
ZTE is closely investing in AI-native community structure and high-efficiency computing. “Each digital utility — together with AI — is determined by environment friendly and inexperienced infrastructure,” mentioned Cui.
The ZTE govt additionally detailed the corporate’s built-in strategy throughout networks, computing and vitality. The Chinese language vendor is deploying modular, prefabricated knowledge middle items and AI-based energy administration, which she mentioned cut back vitality use and cooling hundreds by greater than 10%. These developments are aimed toward delivering versatile, sustainable capability to fulfill rising AI calls for, the chief mentioned.
ZTE can also be advancing “AI-native” networks that shift from conventional architectures to heterogeneous computing platforms, with embedded AI capabilities. This, Cui mentioned, marks a shift from AI as a help software to autonomous brokers shaping operations.
Cui emphasised the function of high-quality, safe knowledge and environment friendly algorithms in constructing extra succesful AI. “Information is like fertile ‘soil’. Its quantity, purity and safety resolve how properly AI as a plant can develop,” she mentioned.
ZTE’s full-stack clever computing options are being utilized throughout client gadgets, good houses and industrial sectors. The corporate is integrating AI into smartphones, PCs, house networks and robots, aiming to create a seamless “perception-computing-interaction-execution” loop in house environments. In industrial use instances, its Nebula fashions have shortened manufacturing cycles and improved vitality effectivity, with deployment throughout greater than 100 tasks in 18 verticals.
“Within the community discipline, we’re driving an architectural revolution with built-in intelligence. This takes community structure from conventional devoted {hardware} to heterogeneous computing platforms, reaching a shift from ‘cloud native’ to ‘AI native.’ That’s greater than only a increase in effectivity — it’s a change in roles,” the chief mentioned. “We’re advancing the evolution of AI from being a mere copilot to an agent that may make choices autonomously. By doing so, we goal to reshape community operations and repair fashions, and unlock new enterprise fashions for clever community functions,” she added.
Throughout MWC Shanghai 2025, ZTE demonstrated a suite of merchandise and options for radio entry community. These developments are designed to empower international operators in deploying 5G networks that includes full-band protection, streamlined website deployment, premium person experiences and deep AI integration, the seller mentioned.
“Responding to surging uplink bandwidth calls for from AI terminals, autonomous driving, and embodied intelligence functions, ZTE’s uplink 3-carrier aggregation (3CC CA) know-how unlocks spectrum potential,” the seller mentioned in a launch. In the course of the occasion, ZTE and China Cell demonstrated this answer on industrial mid-band, reaching uplink speeds exceeding 865Mbps.
“The deep convergence of cell community and AI is pioneering new service paradigms. ZTE’s AIR RAN answer leverages heterogeneous computing, with the AIREngine — a BBU-embedded computing board — as its clever core. By integrating digital upgrades throughout antenna, transmission, and energy programs, it transforms legacy RAN into an AI and RAN deeply built-in clever infrastructure,” the seller added.
