US Data Center Leasing Surges to 7.4GW in Q3 2025, Led by Oracle and OpenAI

Nov 10, 2025
US hyperscalers leased a record 7.4GW of data center capacity in the third quarter of 2025, surpassing the total for all of 2024, according to TD Cowen. Oracle accounted for most of the activity, securing 5.4GW of capacity to support OpenAI workloads.

US hyperscalers leased a record 7.4 gigawatts (GW) of data center capacity in the third quarter of 2025, surpassing the total for all of 2024, according to a report by TD Cowen. The firm said the active US leasing pipeline reached roughly 10.2GW, bringing the year-to-date total to 11.3GW compared to 7GW for the whole of 2024.

Oracle Corporation accounted for the majority of the leasing, securing 5.4GW across multiple US sites, most of which will support workloads from OpenAI. Other major participants included Alphabet subsidiary Google, Meta, Microsoft, Amazon, and Anthropic, each negotiating or signing gigawatt-scale projects.

TD Cowen analysts described the quarter as the largest single increase in data center demand since the industry's inception. Google leased around 600 megawatts (MW) of capacity, while Anthropic followed with 528MW. The surge follows earlier signs of market recovery after a slowdown in early 2025, when Amazon and Microsoft postponed or cancelled several projects.

Meanwhile, Digital Realty reported in a press release that it signed total bookings expected to generate $201 million in annualized rental revenue in the same quarter, reflecting strong enterprise demand and an expanding backlog of $852 million in annualized base rent.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.

Also, consider following us on social media:

AI Chips & Datacenters AI Brief AI Brief (X)
Free newsletter

Silicon Brief

Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.

AI Token Webinar

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You may also like

Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.