Thanks to generative AI, hyperscaler data center capacity will triple by 2030

According to Synergy Research Group, the average capacity of hyperscaler data centers that will launch over the next four years will be almost double the capacity of existing facilities. The critical IT workload of hyperscaler data centers was already growing, but energy-intensive generative AI technologies and services have accelerated this trend.

As the average capacity of individual data centers increases, the number of operating hyperscaler data centers will also continue to grow steadily. There will also be some modernization of existing data centers to increase their capacity. As a result, the capacity of all existing hyperscaler data centers will almost triple by 2030.

The study is based on an analysis of the data center footprint and operations of 19 of the world’s largest cloud and Internet services companies, which Synergy classifies as hyperscalers. This includes the largest SaaS, IaaS, PaaS operators, search engines, social networks, e-commerce services and gaming services.

Image source: Synergy Research Group

By the end of 2024, these companies had 1,103 operational data centers around the world. The list of future data centers includes another 497 objects taken into account in the Synergy forecast model. “The number of operating hyperscaler data centers continues to grow inexorably, doubling over the past five years,” said John Dinsdale, principal analyst at Synergy Research Group, adding that GPU-centric infrastructure will lead to a doubling of the capacity of new hyperscaler data centers.