The Nuclear Summit at Datacloud Global Congress 2026 brought together data centre operators, energy providers, and nuclear technology developers to examine whether the sector’s most reliable low-carbon energy source can scale fast enough to meet the industry’s power demand. The finding was cautiously optimistic-with two significant caveats: community acceptance is the primary barrier, and the fuel supply chain cannot be assumed.

The data centre industry’s power problem has produced an unlikely alliance. Operators who spent a decade building procurement strategies around renewable energy power purchase agreements (PPAs) are now actively exploring nuclear as a source of baseload power that renewables cannot provide. Wind and solar are variable. AI inference workloads are persistent and power-dense.

The combination creates a firm power requirement that renewables, without significant firming arrangements and storage, cannot reliably meet. Nuclear-particularly advanced reactors and small modular reactors (SMRs)-offers what the AI era demands: always-on, low-carbon, dense power generation. The question is not whether the physics work. It is whether the economics, the communities, and the regulatory environments can support deployment at the scale and speed the data centre market requires.

The community acceptance problem is the primary barrier

The nuclear panel at DCGC 2026 was emphatic that community acceptance, not technology, is the binding constraint on nuclear deployment. In the United States, organisations such as Idaho National Lab and TerraPower-backed by Bill Gates-have demonstrated that community acceptance is achievable in the right context.

The example of Kemmerer, Wyoming was discussed at length: a community facing economic decline from a retiring coal plant actively sought to host TerraPower’s first advanced reactor, viewing it as a source of revitalisation and long-term employment. This is a materially different dynamic from the opposition typically faced by data centre projects in established markets.

The panel acknowledged, however, that NIMBYism and misinformation remain significant barriers in many markets, and that in some cases opposition has escalated to extreme actions. In Europe, acceptance varies significantly by country: Sweden, Finland, the UK, and France were cited as markets where nuclear is regaining political and public support, particularly in regions affected by industrial decline.

SMRs, advanced reactors, and the technology pipeline

The technology landscape for nuclear has evolved considerably in the past five years. TerraPower’s next-generation reactor and the broader pipeline of SMR designs offer characteristics that make them attractive for data centre co-location: modular deployment, flexible siting requirements, and lower upfront capital requirements relative to large-scale conventional reactors.

The integration of nuclear with AI data centre workloads does, however, require sophisticated load balancing, energy storage, and grid management capabilities. The panel noted that nuclear alone cannot meet all the energy needs of a large AI campus-hybrid solutions combining nuclear baseload with gas peaking capacity, battery storage, and grid connections are necessary to handle the highly variable demand profile of AI inference workloads.

Bouke van der Weerdt, Chief Technology Officer at Huawei Digital Power, characterised energy storage systems as indispensable multi-function tools for precisely this reason-capable of filtering the rapid power swings that AI workloads generate and preventing grid-level instability.

Financial structures and fuel supply: the two remaining constraints

The panel addressed the financial and supply chain dimensions of nuclear deployment with candour. In the US, private investment in advanced nuclear projects-often in partnership with data centre operators-is growing, with risk allocation increasingly negotiated between governments, technology vendors, and off-takers on a project-specific basis.

In Europe, projects still depend heavily on government support or favourable regulatory classification, though there is a push toward private-led financing as technology matures. Fuel supply chain constraints received particular attention. The US is investing in new uranium enrichment and fuel recycling infrastructure to reduce dependence on foreign sources, but Europe, outside France, was characterised as lagging on this dimension.

For SMR deployment to scale, fuel availability and processing capacity must be addressed as a strategic supply chain problem, not assumed. On cost, the panel noted that while first-of-a-kind nuclear projects carry a high price, data centre operators increasingly frame the question differently: what is the cost of not having reliable, firm power? For AI infrastructure at scale, that cost is high enough to make nuclear economically rational even at current pricing.

Key Takeaways

• Community acceptance, not technology, is the binding constraint on nuclear deployment for data centre power-and the Kemmerer, Wyoming model shows that industrial communities actively seeking economic revitalisation can be nuclear’s most receptive hosts.
• SMRs and advanced reactors offer modular, flexibly-sited baseload power that addresses the AI infrastructure sector’s firm power requirement-but hybrid solutions with storage and grid backup remain necessary to manage AI’s variable load profile.
• Fuel supply chain constraints-particularly outside France in Europe-are a strategic risk for SMR scaling that requires proactive investment, not assumption.
• Private investment in advanced nuclear is growing in the US, driven partly by data centre operators for whom the cost of power uncertainty exceeds the cost of nuclear investment; Europe is moving in the same direction but from a lower base.
• Water usage concerns around nuclear are largely addressable: many advanced reactor designs use closed-loop or alternative coolant systems with minimal water consumption, contrary to public perception.

Nuclear’s relationship with the data centre sector is at an early but commercially significant stage. The operators making the most progress are those treating nuclear not as a future option but as a current planning input-mapping fuel supply chains, engaging local communities early, and building the financial structures needed to take an equity or off-take position in advanced reactor projects.

The 12 to 24 months following DCGC 2026 are likely to see the first commercial announcements that move this relationship from exploratory to contractual, with the US market leading and Europe following at its characteristically deliberate pace.