Dive Brief:
- Data center spending is on the rise as enterprises accelerate AI adoption and cloud providers race to increase capacity, according to Dell’Oro Group.
- Capital expenditures on cloud infrastructure grew 38% year over year during the first six months of 2024, fueled by an ongoing hyperscaler building boom, the market research firm said. Server and storage system component revenues reached record highs, driven by component cost increases and growing demand for AI-optimized servers.
- As the AI building boom continues, spending on infrastructure is expected to surpass $400 billion this year. Enterprises and secondary cloud service vendors are contributing to the growth, the firm said.
Dive Insight:
Behind every generative AI model release and upgrade there’s a cloud-based data center running to train workloads. As enterprises bring use cases into production, cloud delivers additional data and compute power.
Demand for capacity is reshaping the cloud business.
Massive infrastructure investments upended the balance between two major market segments — data centers and cloud services — during the first half of the year, according to a Synergy Research Group report published last week.
Spending on public and private cloud data centers grew 30% year over year, surpassing a healthy 21% increase in cloud infrastructure, platform and software services revenue, SRG found. The cloud services category is roughly twice as big as the data center market.
“The market for building and equipping cloud-related data centers has received a huge boost from products required to support AI functionality, and growth rates there have for once surpassed cloud service growth rates,” John Dinsdale, a chief analyst at SRG, said in an email.
“Looking ahead five years, service growth will once again outpace the equipment market, but in the meanwhile the vendor landscape will be radically changed, with more tech being sold directly to cloud providers and enterprises, bypassing more traditional system vendors,” Dinsdale said.
The three largest hyperscalers — AWS, Microsoft and Google Cloud — are banking on AI workloads to accelerate enterprise cloud adoption and help fuel massive buildouts, like the $3.3 billion Google committed to new South Carolina facilities Thursday.
Microsoft was busy last week, too. The tech giant announced a $2.7 billion investment in Brazil and committed $1.3 billion to grow its cloud empire in Mexico.
Overall, hyperscaler compute capacity grew by 24% year over year during H1, according to SRG. During the same period, the pipeline of future data center capacity spiked by 47%. The number of hyperscale data centers surpassed 1,000 earlier this year, as cloud compute capacity eclipsed on-prem, according to an August SRG report.
