ZTE Corporation (0763.HK / 000063.SZ), a global leading provider of information and communication technology solutions, announced today it will showcase intelligent computing solutions at the Mobile World Congress 2024. This initiative aims to accelerate digital and intelligent transformation across diverse industries.

Full-Stack Computing Infrastructure Enables Full-Scenario Innovation

Full-stack computing infrastructure offers diversified computing power and high energy efficiency, providing end-to-end solutions for heterogeneous computing power, high-performance storage, lossless networks and full liquid cooling data centers. It meets the requirements of general computing and intelligent computing scenarios.

Oriented to the intelligent computing scenario, the large-scale GPU training pool solution and AiCube can meet the deployment requirements of the 10k+ GPU pool in central DC, heterogeneous computing inference pool in regional DC, and all-in-one training/inference solution at the edge, thus to accelerate the digital and intelligent transformation of the entire industry.

AiCube All-in-One Cabinet for Full-Stack Intelligent Computing

The AiCube is a one-stop compact solution with AI computing, storage, network, and AI platform software. It supports mainstream AI frameworks and meets the requirements for on-premises deployment of edge nodes. It helps enterprises reduce the training and inference costs of private field AI models, lower technical barriers, and ensure enterprise data security.

IceCube High-Density Full Liquid-Cooled Cabinet, Achieving pPUE<1.1

The IceCube is a cutting-edge pre-integrated liquid cooling cabin solution. It can accommodate up to 40 1U servers, providing ultra-high computing density and a smaller footprint. The cabin features a blind mating design, and with liquid cooling cabin doors, it can achieve a pPUE<1.1, making it ideal for building efficient, green data centers.

AI Booster Intelligent Computing Platform Enables Industry AI Innovation

ZTE developed the AI Booster Intelligent Computing Platform to maximize GPU utilization through automatic parallel training and significantly reduce development barriers through visual development, zero-code design, and self-adaptive parameter optimization.

Apply Generative AI to All Scenarios, Facilitating Industry Efficiency Improvement and Modernization

R&D: Develop a coding assistant application to improve overall research and development efficiency by 20% using the large model of Nebula Encoding.

Telecom Field: Enable the commercial use of the autonomous network and ensure the industry’s first smart event based on the Nebula telecommunication large model.

Industrial Park: Implement the Nebula CV model in the campus to support multiple security operation scenarios, increasing overall operation efficiency by 35%.

Find out more about ZTE here.