Hyperscaler data center buildouts and upgrades will nearly triple cloud compute and storage capacity in the next six years, driven by an anticipated jump in high-intensity generative AI workloads, according to Synergy Research Group (SRG).
The research firm analyzed the data center footprint and operations of 19 global cloud and internet service firms, including AWS, Microsoft Azure, and Google Cloud Platform. The report found that hyperscalers operated 926 major data centers globally in mid-2023, with plans underway to open 427 additional facilities. The number has doubled in just the last five years.
“We are seeing some changes to plans for future deployment that are being driven by generative AI technology and services,” John Dinsdale, chief analyst at SRG, said in the report. “These changes are not so much to do with the number of new data centers that will be launched, but rather the capacity and power density of those facilities, as GPUs are deployed in ever-greater numbers.”
Generative AI workloads, such as large language models and image generation tools, require significantly more compute power than traditional workloads. To meet this demand, hyperscalers are investing in new data center designs that can accommodate more powerful servers and other AI-optimized hardware.
For example, Oracle broke ground on six new data centers this year, all of which are designed to support AI workloads. The company also brokered a deal with Microsoft to install its database hardware in Azure data centers.
The proliferation of cloud, data-intensive applications, and digital technologies is driving exponential growth in the data center industry. Emerging technologies, such as generative AI, are adding momentum.
The anticipated swell in generative AI workloads has intensified competition among the three largest hyperscalers to capture market share. AWS, Microsoft, and Google Cloud have all pledged to upgrade infrastructure to meet customers’ AI compute needs.
The sources for this piece include an article in CIODIVE.
