Login

Cloud Computing Firms Increase Data Center Investment for Artificial Intelligence

Google, Meta, Amazon and Others Drive Record Demand, Try to Deal With Power Constraints

Top U.S. data center users including Amazon, Meta and Google have been increasing capital spending to boost cloud computing capabilities, according to Newmark's latest data center report. (CoStar)
Top U.S. data center users including Amazon, Meta and Google have been increasing capital spending to boost cloud computing capabilities, according to Newmark's latest data center report. (CoStar)

The nation's biggest cloud service providers have been increasing their investments to support surging growth in the use of artificial intelligence, leading to record demand for data centers while the sector deals with land and power constraints.

Google, Meta, Amazon, Microsoft and Oracle collectively spent nearly $160 billion in 2022 to boost their cloud computing capabilities, up an average of 30% over the past five years, according to Newmark's U.S. data center market overview report that published in January. That spending trajectory — an indicator of data center demand — is likely to continue because of the accelerating pace of AI breakthroughs, the brokerage found.

"Right now, the demand is much greater than the developers can build, and the driving component is the utility or the power that needs to be brought in," said Bo Bond, a Dallas-based executive managing director in brokerage Cushman & Wakefield's global data center advisory group. "The utility is the driving force behind the lack of supply."

article
3 Min Read
January 30, 2024 06:37 PM
Alphabet has spent more than $1.8 billion to get out of office deals and trim its global workforce.
Katie Burke
Katie Burke

Social

The U.S. data center market had record demand for space in the past year, with more of that industrial property expected to be needed as cloud computing companies sign leases even before projects are completed. That pre-leasing is a rarity in the data center world, Bond said, but has become a part of the business as businesses look to secure prime locations and ensure bandwidth for future growth.

"We were not talking about machine learning at this scale even five years ago," he added. "The usage of the cloud on a daily basis is growing at an incredible rate and they need to house this data to meet the demand and computing power needed to deliver applications."

Bond said data center development must have access to plenty of power, something that may be challenging because, as JLL's Data Centers 2024 Global Outlook found, the cost and availability of power needed to support advanced AI workloads is expected to hinder data center development. Top users have looked at low-carbon strategies to overcome supply issues, including expanding into secondary and tertiary markets, using advanced nuclear reactors for on-site power and power-purchase agreements for renewable energy.