Dive Brief:
- Big tech companies are on pace to pour more than $180 billion into data center expansions and related infrastructure costs, according to Dell’Oro Group. The market research firm shared its analysis of data center capital expenditures with CIO Dive ahead of a December report.
- Amazon, Google, Meta and Microsoft together increased infrastructure investments by 81% year over year during the third quarter of 2024, the firm estimated.
- The building boom was fueled by AI model training, the primary focus of data center investments this year, Dell’Oro Group Senior Research Director Baron Fung said. AWS, Azure and Google Cloud each reported major year-over-year spikes in capital expenditures and AI revenues in Q3 and forecast more of the same in the coming months.
Dive Insight:
Cloud providers invested in AI-optimized infrastructure to supply model builders with compute and prepare for an even larger wave of processing demand. The three largest public cloud providers — AWS, Azure and Google Cloud — accounted for roughly 80% of Q3 spending, Dell’Oro Group found.
While AI chipmaker Nvidia saw revenues nearly double year over year in the three-month period ending Oct. 27, the market for traditional central processing has picked up, too. CPU, memory and storage vendor revenues increased more than 90% in Q3, marking the fourth consecutive quarter of double-digit growth, according to Dell’Oro Group.
Enterprise AI adoption remains in the early stages but, as use cases scale, the research firm expects AI-focused servers to account for a quarter of server revenue in the coming months, Fung said in an email.
The data center component industry enjoyed record returns during Q2, with server revenues growing 127% year over year to $54 billion, according to a September Dell’Oro Group report. Hyperscaler hunger for graphics processing units, coupled with pent-up demand for CPU servers, drove the unprecedented growth.
“General-purpose computing and storage are recovering, particularly at Amazon and Microsoft, as they expand capacity to meet growing cloud service demands,” Fung said.
As AI ripples through the tech sector, network equipment and liquid cooling solutions to support inference and tuning operations will be the next segments to see investments spike.
“Investments are increasing in specialized networking to interconnect AI clusters,” Fung said.
