PRESS RELEASE: In the era of digital economy, all industries are experiencing an unprecedented wave of transformation.

With the rapid development of technologies such as AI, big data, and cloud computing, there is a surge in demand for massive computing power. Traditional architectures are struggling to meet the increasingly complex needs of business scenarios. ZTE has introduced a full-stack intelligent computing infrastructure to serve as a super engine for digital transformation across all industries.

To fulfil the requirement of AI era, the full-stack intelligent computing infrastructure offers a full range of servers, lossless networks, and end-to-end solutions such as AiCube and IceCube cabinet system solution. It provides customers with customizable computing power according to their needs, offering users more personalized choices. From intelligent computing centers  of large model training to hybrid intelligent computing centers of training and inference, and then to AiCube edge training and inference integrated cabinet solution, and IceCube high-density fully liquid-cooled whole cabinet solutions, it meets the differentiated demands for performance, cost, and service in various scenarios.

  • Extreme quality: Through the smart brain, zero errors in the production process is achieved; through six-dimensional force sensors, zero deviation in assembly is achieved; through CV large model detection, zero missed inspections is achieved. By these methods, high-quality server products are produced.
  • Extreme energy efficiency: The entire series of servers supports cold plate liquid cooling and immersion liquid cooling. They adopt vacuum chamber heat spreader plates combined with cold plates and multi-hole fin heat sinks to enhance heat exchange capabilities and reduce overall energy consumption.
  • Extreme computing efficiency: By GPU open interconnectivity and using 51.2T switch chips, high-speed interconnectivity between chips within the server is achieved; through end-network collaboration, high-speed interconnectivity between servers is achieved, constructing a cluster of tens of thousands of cards.

In addition, ZTE has launched complete cabinet solutions that integrate the latest technology flexibly with the fastest speed, achieving higher efficiency:

– AiCube training and inference integrated cabinet integrates computing, storage, network equipment, and AI platform software. It supports mainstream large language models (LLMs) such as DeepSeek, Qwen, and Llama, enabling one-stop deployment and plug-and-play, Lowering the barrier to entry for AI applications..

– IceCube high-density fully liquid-cooled cabinet solution allows for zero-gap efficient deployment of servers, saving more space, achieving higher energy efficiency, and simplifying deployment.

ZTE full-stack intelligent computing infrastructure is not only a powerful tool for enterprises to enhance operational efficiency and competitiveness but also a super engine that drives the comprehensive upgrade of the industry ecosystem. ZTE will continue to deepen its exploration in the field of intelligent computing and engage in in-depth cooperation with global operators and industry partners. ZTE will empower more enterprises to leverage this super engine to jointly promote the advancement of AI technology and digital intelligence upgrades across various industries, opening up new business models.