AI Data-Center Power Demand Highlights Growing Market for On-Site Energy and Electrical Infrastructure

The review frames AI data-center expansion primarily as an energy-infrastructure demand driver rather than as a specific data-center development event. It does not report a new campus, commissioned facility, announced MW capacity or project-level CAPEX.

According to the review’s citation of the International Energy Agency, AI electricity demand is projected to double between 2024 and 2030 and reach 945 TWh by 2030. The article also references 15% annual consumption growth, although it does not provide the methodology or a clearly defined calculation base for that figure.

The investment case presented centers on suppliers spanning distributed power, electrical equipment and thermal infrastructure. The review includes Bloom Energy, Plug Power, Eaton, Vertiv, GE Vernova and NextEra Energy. It describes factory-built fuel-cell systems located near data centers as a potentially faster option for local power supply where grid capacity or connection timing is constrained.

Bloom Energy reported that demand for its fuel cells had increased 140% year on year to $6 billion at the start of 2026, according to the article. This is a demand indicator, not confirmed revenue, deliveries or commissioned generation capacity. Eaton, meanwhile, estimated at the end of the second quarter of 2026 that AI data centers would become its largest end market during 2026 and account for 21% of sales; this is an outlook for the year rather than a completed-period result.

For operators and investors, the underlying implication is that AI data-center economics and delivery schedules could become increasingly dependent on available power and the speed of grid interconnection. That could favor vendors of on-site generation, power distribution, backup-power systems and cooling. However, the review is an investment-oriented market assessment and does not confirm orders, deployments or operational data-center capacity.