The article explains that the rapid growth of artificial intelligence is creating an enormous surge in electricity demand—so much so that traditional energy sources may struggle to keep up. According to analysts at Bernstein, small modular reactors (SMRs) are emerging as a promising solution to power AI-driven data centers. These next-generation nuclear systems offer a stable, continuous energy supply, which is critical for running energy-intensive AI workloads around the clock.
A key advantage of SMRs is their scalability and potential cost efficiency. Bernstein estimates that, once deployed at scale, SMRs could reduce costs by up to 70%, making them competitive with existing energy sources. Unlike traditional nuclear plants, which are large and expensive to build, SMRs are designed to be smaller, modular, and easier to deploy—making them particularly suitable for colocating near data centers that require reliable, high-density power.
The growth projections for this sector are significant. Bernstein forecasts that global SMR capacity could rise from about 280 megawatts in 2025 to 4.2 gigawatts by 2035, representing a strong expansion trajectory. While this would still represent a small share of total global nuclear capacity, the increasing demand from AI infrastructure is expected to accelerate adoption, especially as tech companies look for long-term, stable energy solutions.
Overall, the article highlights a major shift in the relationship between energy and technology. As AI systems become more powerful and widespread, energy infrastructure is becoming a strategic priority. Small nuclear reactors could play a crucial role in this transition—positioning nuclear energy not just as a clean power source, but as a key enabler of the AI-driven digital economy.
