In partnership with
France's 56 operational reactors sit within 50 kilometres of more than 40 potential datacenter locations. China's 32 reactors under construction are creating new power corridors that hyperscalers have already begun mapping. Across 31 countries, our community has now identified over 200 sites where nuclear capacity and datacenter demand converge. The question is no longer whether nuclear powers AI. It is who maps the opportunity first.
440 reactors, zero commercial maps
The IAEA's PRIS database tracks 440 operable reactors across 31 countries. The World Nuclear Association confirms 70 more under construction. Yet no single public resource maps where these facilities intersect with datacenter demand corridors, grid capacity, and land availability.
Hyperscalers are solving this privately. Microsoft's £12.6bn ($16bn, €15.1bn) Three Mile Island restart targets 835 MW by 2028. Meta's 20-year agreement with Constellation secures 1,121 MW from the Clinton Clean Energy Centre in Illinois. Amazon's Susquehanna campus could draw over 1.9 GW. Google signed with Kairos Power for 500 MW of Small Modular Reactor (SMR) output by 2035.
These are billion-pound decisions made with proprietary intelligence. Smaller operators, investors, and governments lack equivalent data. That asymmetry costs the market billions in missed opportunities.
£50,000 reports that arrive too late
Government databases track reactors but ignore commercial potential. The IAEA catalogues technical specifications, not proximity to fibre routes or grid headroom. National regulators publish safety data, not investment signals.
Commercial real estate platforms map datacenters but treat power as a commodity input. They model megawatts available, not megawatts reliable. Nuclear's 90%+ capacity factor, the percentage of maximum possible output actually delivered, rarely features in site selection models.
Industry consultancies sell bespoke reports at £50,000 ($63,500, €60,000) per engagement. By the time the analysis arrives, the best sites are under letter of intent. Speed kills in this market. The US Department of Energy identified 16 federal sites for rapid datacenter development. Within weeks, most had received commercial expressions of interest.
The community data flywheel
Our approach inverts the traditional model. Instead of top-down analysis, we built a contributor network across four continents. The results speak through the people who built them.
The methodology works in three layers. First, IAEA PRIS data provides the foundation: reactor locations, capacity, operational status, and grid connections. Second, contributors add local intelligence: land availability within 50 kilometres, fibre route proximity, water cooling access, and planning constraints. Third, satellite monitoring validates claims and tracks construction activity in near real-time.
Consider one example. France's Bugey nuclear site generates 3.6 GW near Lyon. A contributor flagged that the adjacent industrial corridor has fibre connectivity, water cooling access, and grid headroom. That site assessment took days. A consultancy would have billed months for the same insight.
The network effect mirrors what Waze achieved for traffic: each contributor improves the dataset for everyone. Waze reached 50 million users and a £1bn ($1.3bn, €1.2bn) acquisition by Google because crowdsourced data compound in value. Our community operates at a different scale but the same principle. Every verified site makes the next verification faster.
Quality control matters enormously. Contributors must cite primary sources: grid operator filings, planning applications, environmental assessments. We cross-reference against satellite imagery and PRIS operational data. The Alan Turing Institute's research on crowdsourced science confirms that structured validation protocols bring community data to institutional quality.
Community voices reinforce this. Jeff Nolan, an infrastructure standards veteran who serves on international data harmonisation committees, sees the shift clearly: "The best site intelligence comes from people who know the local grid, the local planning authority, and the local terrain. No database captures that." Hong Gao, whose career spans infrastructure interoperability engineering, adds a sharper point: "By the time a traditional report reaches your desk, the data is already stale. Community networks update in real time."
What matters most for nuclear-DC site selection?
Why open beats proprietary
The market values proprietary intelligence. Hyperscalers spend billions securing exclusive access to nuclear sites. Yet the infrastructure investment community, managing trillions in assets, relies on fragmented public data.
This creates a temporal arbitrage opportunity. Communities that pool intelligence move faster than organisations that hoard it. OpenStreetMap proved this for geographic data, reaching completeness that rivals commercial mapping services. Our model applies the same logic to energy infrastructure intelligence.
The window matters. China approved 10 new reactors in 2025 alone. Its Linglong One SMR at Changjiang targets commissioning in early 2026. It would be the world's first commercial onshore SMR. The UAE's Barakah plant delivers four reactors totalling 5.6 GW. Every new reactor creates new datacenter adjacency opportunities that the community can map within days, not quarters.
Governments are catching on
Governments are beginning to recognise the value of open infrastructure intelligence. The US DOE's 16-site identification for datacenter development marks a shift from classified to collaborative. Energy Secretary Chris Wright noted these sites are "uniquely positioned to host data centers as well as power generation."
The UK's approach differs but converges. The Culham AI Growth Zone repurposes a former fusion research site. Great British Energy Nuclear targets Wylfa for SMR deployment. Both signal that governments want private sector intelligence to flow into public infrastructure planning, not away from it.
Three principles, one network
Three principles guide the next phase. First, open intelligence beats closed analysis when speed determines value. The Q2 window closes fast. Second, contributor recognition builds the network. Every verified site carries attribution to the person who mapped it. Jeff, Hong, and dozens of contributors across four continents are named in the dataset they built. Third, regional expertise compounds globally. A contributor who understands French planning regulations adds value to every EDF-adjacent site worldwide.
What this means for your portfolio
For infrastructure investors, community-mapped intelligence compresses site selection from months to queries. The 200 sites already identified represent potential capacity intersections worth tracking.
For hyperscalers, the message is different. Proprietary advantage erodes when communities move faster. The strategic response is engagement, not exclusion.
For governments, crowdsourced infrastructure intelligence offers a model that scales without public expenditure. The community does the mapping. Policy frameworks enable the building.
Bottom line
Two hundred sites mapped. Four continents covered. Thirty-one countries represented. Not built by a consultancy billing £50,000 ($63,500, €60,000) per report. Built by Jeff, Hong, and a growing community who understood what the market missed. In infrastructure intelligence, the network is the product. The window to join it is open now.
Next week: China built 1.2 GW of nuclear-AI capacity in two years. We examine how, and what the rest of the world can learn.
Why Burned-Out Professionals Are Turning to CBD
Long hours, heavy workloads and frustrating coworkers: job burnout is on the rise. Savvy professionals are turning to CBD and its sister compound, CBG, to manage work stress and burnout without losing focus.
Like CBD, CBG is a non-intoxicating compound found naturally in the hemp plant that is known for helping to manage mild/temporary anxiety. And these benefits are backed by science: participants in a 2022 CBDistillery study who took CBD + CBG reported:
Reduced mild or temporary anxiety and tension
Increased focus and mental clarity
That CBD + CBG was more effective than other treatments they tried in the past
When it comes to quality CBD products, CBDistillery is leading the way. Non-GMO, US-grown hemp, 100% clean ingredients, and a 60-day satisfaction guarantee give you one less thing to stress about.
Special Deal for Our Readers: Use code FOCUS to save an additional 20% on your first bundle order!
