Alibaba Cloud Plans to More Than Double Global Modular Data Center Capacity This Year

AI Author: EqualOcean News Updated 7 hours ago (GMT+8)

Alibaba Cloud (阿里云) said on August 10 that it plans to more than double its global modular data center capacity this year and can deliver large AI data-center projects in about 100 days using its modular infrastructure approach.

black tech

The company did not disclose a capacity baseline or a target expressed in megawatts. Its announcement positions the expansion as a response to rising demand for AI computing infrastructure, where delivery time, power systems and cooling capacity have become central constraints on new deployments.

Alibaba Cloud said its modular approach allows key systems to be manufactured and prepared in parallel rather than installed sequentially on site. The architecture divides power, cooling, security, intelligent-management and fire-protection systems into prefabricated modules that can be assembled at a project site.

According to the company, its CUBE DC 5.0 architecture raises the modularization rate of those five systems from about 30% to 90%. Alibaba Cloud said a typical 100-day delivery schedule comprises 30 days for factory production and site preparation, 50 days for on-site installation and 20 days for commissioning.

The company also said the modular approach can reduce overall construction costs by more than 10%, although the outcome will depend on project scale, site conditions, power availability and local regulatory requirements.

CUBE DC 5.0 was introduced by Alibaba Cloud in 2024 as a proprietary architecture for high-density and energy-efficient data centers. At the time, Alibaba Cloud said the design could support up to 200MW of IT capacity and rack densities of up to 200kW at full data-center build-out.

For Alibaba Cloud, the announcement is primarily a capacity-deployment plan rather than evidence that the additional infrastructure has already entered service. The key next indicators will be the locations of the planned deployments, their commissioned IT capacity, and whether the 100-day timeline is achieved in overseas projects as well as in China.