Dell'Oro lifts data centre capex forecast on AI demand

Dell'Oro Group has forecast that worldwide data centre capital expenditure will surpass $3 trillion by 2030, nearly double its January 2026 outlook.

The revised forecast points to sustained spending on AI infrastructure, with high-end accelerators used in AI-optimised servers expected to take the largest share of data centre investment through the end of the decade. Those systems are also expected to remain the main driver of expenditure growth.

The change reflects a broader reassessment of how much the largest cloud operators are likely to spend on data centres and associated equipment. It also incorporates higher estimates for global data centre power capacity and rising commodity costs.

Baron Fung, Vice President of Research at Dell'Oro Group, outlined the factors behind the revision.

"Our 2030 data center capex outlook has nearly doubled since the January 2026 forecast, reflecting higher hyperscale capex guidance, increased projections for global data center power capacity, and higher commodity costs," Fung said.

The shift comes as investors and suppliers watch whether spending on AI computing systems can be sustained at current levels. Data centre expansion has become one of the clearest markers of the race among major cloud groups, chipmakers and newer AI infrastructure providers to secure enough compute and power for training and inference workloads.

The Top 4 US hyperscalers alone could account for about half of global capital expenditure. That concentration underlines how closely the market remains tied to decisions by a small group of companies with the balance sheets to fund large-scale data centre construction and equipment procurement.

AI demand

Spending is expected to remain heavily tilted towards accelerated computing, but demand for general-purpose servers should also benefit from rising inference, agentic AI and storage workloads. That suggests infrastructure growth may extend beyond the most expensive AI systems, even if those machines continue to command the largest share of budgets.

Another projected growth area is the segment Dell'Oro describes as AI-specialised cloud providers, which includes AI model builders and neocloud service providers. This newly added segment is expected to grow at a compound annual growth rate of nearly 60% through 2030, outpacing other customer groups in the report.

The neocloud category has emerged as a distinct part of the market as specialist providers rent access to graphics processing units and related systems to AI developers that do not want to build their own data centre estates. Some of these groups have expanded quickly by targeting demand from model developers and enterprises experimenting with generative AI tools.

Constraints remain

Even with the stronger forecast, Dell'Oro cautioned that growth will depend on several external factors. Access to power is becoming a central issue in many major data centre markets, while supply chains for advanced chips, networking equipment and other infrastructure remain under pressure as demand rises.

Fung addressed those risks in the firm's market commentary.

"High-end accelerators powering AI-optimized servers are expected to account for the largest share of data center capex and remain the primary driver of growth over the forecast period. However, the pace of growth will depend on the sustainability of investment, power availability, and supply chain conditions. Accelerated and heterogeneous computing, along with innovations in server efficiency, could help mitigate the rising cost and infrastructure demands of AI. The Top 4 US hyperscalers alone could represent about half of global capex, while enterprise investment remains constrained by uncertain AI returns," Fung said.

The reference to constrained enterprise investment highlights a split in the market. While hyperscalers and AI-focused cloud providers continue to commit large sums to infrastructure, many corporate users are still assessing whether AI applications will deliver returns that justify major capital spending on in-house systems.

That divide has implications for suppliers across the server, storage, networking and power equipment markets. Vendors exposed to large cloud and AI accounts may benefit from concentrated spending, while those reliant on broader enterprise refresh cycles could face a slower and less predictable recovery.

Dell'Oro's forecast also points to the growing importance of physical infrastructure alongside compute hardware. As AI clusters become larger and denser, operators must invest not only in processors and servers but also in power distribution, cooling and site development to support higher electricity loads.

As a result, growth in data centre spending is no longer defined only by server volumes. It is increasingly tied to a wider build-out of facilities and supporting systems needed to house energy-intensive AI equipment, with the largest cloud operators expected to remain at the centre of that expansion.