Home / Business and Economy / AI Growth Capped by Looming Power Shortage
AI Growth Capped by Looming Power Shortage
12 Jun
Summary
- AI data center expansion faces existential threat from power constraints.
- Half of US data centers may not break ground this year due to power issues.
- Massive water usage by data centers adds to public relations challenges.
The ambitious build-out of artificial intelligence data centers is encountering a significant hurdle: a lack of sufficient power. Energy constraints are posing an existential risk to the expansion plans of AI hyperscalers, with projections indicating that half of the planned data centers in the US might not begin construction this year, and more may face delays in 2027. This situation has considerable implications for the broader economy, as AI infrastructure development has been a key driver of GDP growth.
Companies are struggling to secure the necessary power, facing resistance from traditional grid suppliers and challenges in developing their own power sources quickly enough. While nuclear power is seen as a long-term solution, its development is a lengthy process. Adding to the complexity, data centers are also extremely water-intensive. Estimates suggest a hyper-scale data center can consume between 1 to 5 million gallons of water daily, raising further concerns about resource availability and public relations.
