Neil Chatterjee, the chief of government affairs at Palmetto, said nuclear power should support the electricity needs of data centers.
This proposal comes as artificial intelligence drives a surge in electricity demand, forcing energy providers to find reliable sources that can maintain constant operations without increasing carbon emissions.
Speaking during an interview on CNBC’s ‘Power Lunch,’ Chatterjee said that AI-driven data centers require a steady supply of low-carbon electricity. He said that nuclear energy is uniquely positioned to provide this carbon-free supply to meet the rigorous demands of the tech sector.
"Nuclear power should support the power needs of data centers," Chatterjee said.
The scale of the energy challenge is significant. U.S. data-center electricity consumption is projected to rise from 17 GW in 2022 to 35 GW by 2030 [1]. This growth creates a gap that traditional renewable sources, such as wind and solar, may struggle to fill alone due to their intermittent nature.
However, the viability of nuclear energy as a primary solution is a point of industry debate. Some analysts said nuclear power and small modular reactors are the primary solution to data-center energy woes [2]. Others said nuclear power and these smaller reactors will be unable to fill the electricity needs of data centers, particularly in the near term [3].
Chatterjee's advocacy reflects a broader push to integrate baseline nuclear power into the digital infrastructure of the U.S. to ensure that the expansion of AI does not compromise climate goals.
“"Nuclear power should support the power needs of data centers."”
The tension between the energy-intensive nature of AI and corporate net-zero pledges is creating a strategic pivot toward nuclear energy. While the projected increase in power demand is stark, the disagreement between analysts regarding the timeline for small modular reactors suggests that the transition to nuclear-backed data centers may face significant deployment delays despite political and corporate support.
