AI's Soaring Demand: Data Centers to See 220% Power Surge by 2030, Goldman Sachs Predicts

A new analysis from Goldman Sachs reveals a significant upward revision in projections for global data center electricity usage. The primary catalyst for this unprecedented surge is the rapid and extensive deployment of artificial intelligence infrastructure worldwide. By the close of the decade, data center power consumption is expected to climb by an astonishing 220% from its 2023 levels.

Report Highlights: Data Center Power Demand to Quadruple by 2030

Published on April 14, 2026, the report from Goldman Sachs, highlighted by financial commentary platform The Kobeissi Letter, projects that global data center electricity demand will escalate by 905 terawatt-hours (TWh), reaching a colossal 1,350 TWh by 2030. This updated forecast represents a substantial increase from earlier estimates, which anticipated a 175% growth rate. The acceleration is attributed to a higher-than-expected shipment of AI servers and a widespread industry shift towards energy-intensive hardware essential for advanced AI processing. The United States is poised to bear the brunt of this energy expansion, with approximately 60% of the new global power demand originating from within its borders. This marks an increase from previous forecasts of around 50%. Consequently, U.S. data center power requirements alone are predicted to hit approximately 750 TWh by 2030, with the remaining 600 TWh distributed across other nations. To accommodate this immense need, the domestic data center capacity is projected to soar by 197% between 2025 and 2030, reaching an impressive 95 gigawatts. This escalating demand is already exerting considerable pressure on existing power grids. Data centers currently account for roughly 6% of the total U.S. electricity demand, a figure that industry experts predict could rise to 11% by 2030. As major technology companies continue to rapidly expand their computing capacities, regional power markets are nearing critical limits. This impending energy crisis is compelling a reevaluation of strategies for utility and energy infrastructure investments, as securing dependable power supplies becomes the most pressing challenge in the global AI race. The increased data center electricity consumption is expected to redefine U.S. power markets, offering substantial opportunities for utility companies like Constellation Energy Corp. and Duke Energy Corp., as well as grid infrastructure providers such as Quanta Services Inc., Eaton Corp. PLC, and Bloom Energy Corp. Major tech giants like Alphabet Inc., Amazon.com Inc., and Microsoft Corp., while facing increased energy capital expenditures, are well-positioned due to their vast resources and strong financial standings to navigate and capitalize on this evolving landscape.

This analysis underscores the transformative impact of AI on global energy consumption and infrastructure. The burgeoning demand for data center power necessitates immediate and strategic investments in robust energy solutions. As the United States emerges as a central hub for this expansion, it faces both significant challenges in maintaining grid stability and unparalleled opportunities for innovation in energy generation and distribution. Addressing these energy demands will be crucial for sustaining technological progress and economic growth in the AI era.