Can nuclear fuel the rise of AI?

The escalating power demands from AI and data centers are positioning Small Modular Reactors (SMRs) as a critical, stable, and low-carbon energy solution, with data center electricity consumption projected to double by 2030. Despite significant policy momentum and 28GW of nuclear capacity already announced for U.S. data centers, widespread SMR commercialization faces challenges including cost inflation and regulatory complexity, with most projects not expected until the 2030s. Investors are monitoring key beneficiaries like Doosan Enerbility and Cameco, but tangible progress is required to validate current valuations amid these substantial growth projections.

Analysis

The accelerating power requirements of artificial intelligence and data centers are creating a structural demand for stable, low-carbon energy, positioning nuclear power, particularly Small Modular Reactors (SMRs), as a critical long-term solution. Data center electricity consumption, which stood at 415 terawatt-hours (TWh) in 2024, is projected to more than double by 2030, creating a significant challenge for existing grid infrastructure. SMRs, with their sub-300MW output, faster construction timelines of 3-5 years, and flexible siting capabilities, are well-suited to meet this demand, evidenced by 28GW of nuclear capacity already announced for U.S. data centers. However, significant headwinds temper the outlook. The technology remains nascent, with most projects not expected to reach commercial operation until the 2030s, and early deployments have been plagued by severe cost inflation, with capital expenditures sometimes tripling initial estimates. Despite a projected 31% compound annual growth rate, global SMR capacity is only expected to reach 4.2GW by 2035, representing just 1% of today's total nuclear capacity. Key industry players identified as potential beneficiaries include component manufacturer Doosan Enerbility and uranium producer Cameco (CCJ), which is favored for its low-cost assets and strategic ownership in Westinghouse.

Key Decisions for Investors

• Investors should view the nuclear for AI theme as a long-term, multi-decade investment, given that widespread commercialization of SMRs is not anticipated until the 2030s.
• Carefully monitor execution risk, paying close attention to capital expenditure trends on early projects and milestones toward commercial operation, as these are the primary hurdles to justifying current valuations.
• Consider exposure across the nuclear value chain, from upstream uranium suppliers like Cameco (CCJ) to key component manufacturers, to diversify risk within the sector.
• Track regulatory developments in the U.S. and EU, as policy reforms and capital allocation are significant catalysts for accelerating the deployment and cost-effectiveness of SMR technology.