SHANGHAI, Sept. 19, 2025 -- Huawei Connect 2025 kicked off in Shanghai. At the same time, Huawei Digital Power held the Data Center Infrastructure Summit, where more than 500 business leaders, technical experts, and ecosystem partners from the global data center industry gathered to explore new technologies and practices of green and low-carbon digital infrastructure in the era of industry intelligence. Huawei Digital Power also showcased its green and low-carbon all-scenario digital power solutions and success stories, working with the industry to build high-quality, secure, and reliable data centers in the AI era.

In his speech, Bob He, Vice President of Huawei Digital Power, emphasized that AI DCs are propelling infrastructure toward "computing factories" amid the explosive growth of AI computing power. Efficient and reliable power transmission from grids to chips requires simplifying conversion links and driving innovation in power supply architecture. Thermal management can benefit from a hybrid architecture with adjustable air-liquid ratio to adapt to dynamic AI service requirements. To meet the demands of the AI era, future infrastructure must be built with AI readiness in mind, turning the unpredictability of traditional data center engineering into productized and prefabricated AI DC solutions. Huawei will intensify R&D investment in power supply, thermal management, and infrastructure. Additionally, Huawei provides high-quality products and solutions to boost the construction of reliable, agile, and sustainable AI DCs, and work with the industry to thrive in the golden era of intelligent computing.

Steve Kim, President of Data Center Facility & Critical Power Marketing, Sales and Services, Huawei Digital Power, said at the summit that AI foundation models expedite intelligent transformation of industries and propel the explosion of global GW-scale clusters. The scale of new AI DCs is expected to reach 112 GW in the next five years, predicted by authoritative organizations. In response to the challenges in terms of safety, delivery, IT evolution, and energy consumption in the intelligent computing era, Huawei proposes AI DC construction guideline RAS.

Based on continuous innovations in technology and architecture, Huawei provides comprehensive AI DC solutions featuring all-scenario reference design, highly reliable products & solutions, global ecosystem, and full-lifecycle services, building reliable, agile, and sustainable intelligent computing infrastructure and empowering customers to seize the immense opportunities of the AI era.

During the event, Huawei released the AIDC Facility Reference Design White Paper, which delivers systematic guidance for the standardized design and construction of AI DCs. The White Paper provides an in-depth analysis of challenges faced by high-density cabinets in AI DCs in terms of cooling, power supply and distribution, floor height, and load-bearing capacity, and offers design suggestions across four key dimensions of architectural structure, power supply and distribution, cooling, and integrated cabling, providing multiple reference design solutions for AI DCs based on different power densities.

At the event, Digital Power demonstrated innovative green and low-carbon solutions with a focus on efficient power consumption, electrification, and low-carbon power generation.

With ongoing improvements in computing infrastructure and rapid progress in model deep reasoning and multi-model fusion technologies, application scenarios and value of AI are shifting from quantitative growth to qualitative transformation. By continuously investing in innovation and integrating digital and power electronics technologies, Huawei Digital Power aims to build a robust foundation for intelligent computing that enables more computing power from each watt and foster industry-wide collaboration to build a better, greener future.

Photo - https://mma.prnewswire.com/media/2776637/Vice_President_Huawei_Digital_Power_Bob_He.jpg

Photo - https://mma.prnewswire.com/media/2776638/Huawei_released_AIDC_Facility_Reference_Design_White_Paper.jpg