NTT DATA, a global leader in AI, digital business and technology services, released a global data center outlook report developed in partnership with NTT Global Data Centers and economic consultancy ThoughtLab. The report, titled Can Data Centers Keep Pace with AI? A Global Data Center Outlook, models three expansion scenarios through 2030 and stress-tests likely growth paths to examine whether the infrastructure required to support the next wave of artificial intelligence can scale fast enough to meet demand. The analysis projects annual demand growth of 23% to 30% in the most likely scenarios but warns that capacity constraints across power, equipment supply chains, land availability, and labor could create a capacity crunch if not addressed through coordinated action.
NTT Global Data Centers Report Models Three Growth Scenarios Through 2030
In its stress test of likely growth paths, NTT Global Data Centers found that power availability and grid connections are becoming decisive constraints in major markets, particularly in the United States and Europe. Processors, transformers, switchgear, and backup generators are emerging as significant choke points, with long lead times and limited manufacturing capacity affecting how quickly new projects can be fitted out and energized. Rising community opposition and land constraints are delaying approvals in prime markets, while shortages in specialized construction labor are increasing execution risk and extending delivery timelines. The report argues these constraints are addressable and provides a roadmap for enterprises, operators, investors, and policymakers to unlock capacity and improve the performance economics of AI infrastructure.
Doug Adams, CEO and President, NTT Global Data Centers, Stresses Operational Constraints
"AI demand is accelerating faster than many parts of the underlying infrastructure system can respond," said Doug Adams, CEO and President, NTT Global Data Centers. "The challenge now is not simply scaling capacity, but removing the operational and supply-side constraints that delay deployment and erode the economics of AI investment. This report is intended to help the market move from recognizing the challenges to acting on practical solutions." The recommendations include co-planning power and grid infrastructure with utilities early to align new projects with generation, transmission, storage, and interconnection realities before bottlenecks delay deployment, and strengthening supply-chain resilience by diversifying suppliers, securing longer-term procurement agreements, standardizing equipment specifications, and treating long-lead components as strategic priorities rather than late-stage purchasing decisions.
Report Recommends Performance Benchmarks and Community Engagement Strategies
The outlook calls for driving efficiency-focused innovation through advanced cooling, workload optimization, liquid and direct-to-chip cooling, and AI-enabled operations that reduce pressure on energy and water resources. It also recommends setting stronger and more transparent performance benchmarks, including more consistent use of power usage effectiveness (PUE) and water usage effectiveness (WUE), to improve planning and investor confidence around efficient capacity expansion. Additionally, the report urges improving community engagement and siting strategy so projects can move faster with clearer public understanding of economic benefits, infrastructure impacts, and mitigation measures. "AI infrastructure demand is no longer a future scenario. It is here now," Adams added. "The organizations that move fastest over the next several years will be those that understand where the real constraints are, act early to mitigate them, and build with efficiency, resilience and long-term value creation in mind."
Key Takeaways
- NTT DATA's report projects global data center demand growth of 23% to 30% annually through 2030 across three modeled scenarios.
- Power availability, grid connections, equipment supply chains (processors, transformers, switchgear, generators), land constraints, and specialized labor shortages are identified as primary capacity constraints.
- The report recommends early utility co-planning, supply-chain diversification, efficiency innovation (liquid/direct-to-chip cooling), standardized PUE/WUE benchmarks, and improved community engagement to unlock capacity.
EnergyInsyte's Take
The report quantifies what operators already know: infrastructure lead times are outpacing AI demand signals. The 23–30% annual growth range through 2030 implies a doubling of capacity every 2.5–3 years, which grid interconnection queues and transformer lead times cannot currently support. Buyers and developers should treat the recommended early utility co-planning and long-lead component procurement as immediate procurement priorities, not planning exercises.
Source: Businesswire
