The Impact of AI on Data Center Infrastructure
In this white paper, RSP Principal and Mission Critical Lead Rajan Battish details how AI-driven data centers are transforming the U.S. energy landscape. Once focused on research and enterprise, data centers now represent one of the fastest-growing demands on the electric grid. Goldman Sachs projects AI consumption could reach 500 TWh, with PJM forecasting a 16% increase in demand over the next 15 years. Hyperscale facilities, fueled by investments from major tech firms, refresh hardware every two years and require more power-dense cooling, pushing efficiency metrics higher than traditional centers. While past innovations like virtualization and cloud consolidation kept energy use flat, the scale and speed of AI adoption may overwhelm supply, raising the risk of an “AI bubble” similar to the overbuilding of the Dot Com era.
This paper was first published by Johns Hopkins School of Advanced Studies as a capstone to Mr. Battish’s Master of Arts in Sustainable Energy.
