The rapid evolution of generative AI has unlocked new possibilities across industries, from content creation to advanced data analytics. However, this growth brings an enormous need for data and the infrastructure to process it, resulting in an unprecedented surge in energy demand. As companies push AI capabilities further, they grapple with how to sustainably power this data-driven future.
Generative AI models rely on massive datasets to improve their algorithms, translating into extraordinary requirements for computing power. The Electric Power Research Institute (EPRI) projects that data center energy usage in the US will more than double by 2030, reaching up to 35GW in scenarios with high AI adoption, and that data centers could consume up to 9% of the nation’s electricity. This aligns with broader industry expectations of rapid expansion as AI applications become more computationally intensive.
As energy demands rise, tech companies like Alphabet are exploring nuclear power to support their AI operations. Alphabet recently partnered with Kairos Power to deploy small nuclear reactors for its data centers, with plans for the first reactor to be operational by the end of this decade and additional reactors by 2035. This initiative aligns with Alphabet’s commitment to clean energy and sustainable growth in AI.
This shift reflects a broader trend in tech. Microsoft has secured nuclear energy agreements, and Amazon is exploring similar options for its data centers. By integrating nuclear power, these companies aim to meet AI’s rapidly rising energy needs while advancing environmental goals.
Constellation Energy, the largest nuclear power provider in the US, exemplifies how nuclear energy can support AI’s demands. With 3.5 times the capacity of its nearest competitor, Constellation allows hyperscale data centers to situate near its power sources, reducing transmission time and costs—a significant advantage for companies expanding AI infrastructure.
The rapid growth of generative AI is driving energy demands to new heights, with models like ChatGPT-4 now training on 1.8 trillion parameters and expected to reach tens of trillions by 2030. This surge in data and computational needs is outpacing current infrastructure, pushing companies to find sustainable energy solutions. The future of AI will depend not only on advancements in intelligence but also on developing energy systems capable of supporting these powerful, data-intensive models efficiently.
The information in this article should not be regarded as a description of services provided by Delian Partners SA. The opinions expressed in this article are for general informational purposes only and are not intended to provide specific advice or recommendations for any individual or on any specific security or investment product. It is only intended to provide education about the financial industry. The views reflected in this article are subject to change at any time without notice.
Comments