According to a report by Jefferies, global demand for data centres continues to far exceed available supply, creating a structural capacity shortage that is expected to persist for years as artificial intelligence (AI) adoption accelerates.
The report said that despite aggressive capital expenditure by major technology companies and rapid expansion of AI infrastructure, supply chain constraints are preventing the industry from adding sufficient capacity to meet demand.
As per the report, only 8.9 gigawatts (GW) of capacity became operational in 2025, while demand reached nearly 21.1 GW, resulting in a shortfall of around 12 GW.
The report highlighted a sharp rise in spending by hyperscalers such as Amazon, Microsoft, Google and Meta. Jefferies estimates that hyperscaler capital expenditure will reach $770 billion in calendar year 2026, representing a 74 per cent year-on-year increase and nearly five times the $156 billion invested in 2023.
Further the report mentioned that cloud service backlogs have expanded to around $2 trillion, reflecting sustained demand for computing infrastructure driven by AI applications.
The report also noted that semiconductor shipment forecasts indicate a significant increase in AI-related power demand. Based on projected accelerator chip volumes, global AI data centre power demand could reach approximately 30 GW in 2026, with nearly 19 GW expected in North America alone.
The report identified several supply-side constraints limiting the pace of data centre construction, including shortages of engineering and construction labour, cooling equipment, power infrastructure, transformers, electrical components and backup generators.
According to the report, the ongoing imbalance between demand and supply is benefiting data centre operators, infrastructure developers and equipment suppliers. Vacancy rates remain low across major markets, while leasing activity and rental rates continue to rise.
It further said that the cumulative deficit of undelivered data centre capacity reached 20.4 GW between 2021 and 2025 and is expected to widen further as AI adoption continues to drive demand for computing infrastructure.
The report added that hyperscalers may increasingly secure capacity several years in advance as supply constraints continue to delay project delivery timelines.
