The Uranium Bottleneck and the AI Data Center Grid Shock

About This Report

The convergence of AI-driven data center expansion and advanced nuclear power ambitions has exposed a structural fault line in the uranium fuel supply chain. With data center power demand projected to reach 1,050 TWh annually by 2026 and Goldman Sachs estimating 160–165% growth by 2030, the case for small modular reactors has never appeared more compelling — yet the path to deployment is blocked by bottlenecks few market participants fully appreciate.

Report Sections Overview

• Uranium fuel cycle mechanics — mining, conversion, enrichment, and fabrication
• HALEU enrichment constraints — capacity analysis and scarcity dynamics
• Conversion and fabrication bottlenecks — mid-cycle processing dislocation
• AI data center power demand thesis — 1,050 TWh baseline and growth trajectory
• SMR technology landscape — reactor designs, fuel requirements, and technical readiness
• Realistic deployment timelines — regulatory, construction, and fuel supply pathway analysis 2026–2036
• Key players — Centrus Energy, URENCO, Orano, NuScale Power, TerraPower (Centrus's regulatory moat is the worked example in IU University's Translation Layer guide)
• HALEU economics — enrichment cost structure and commercial viability
• Nuclear fuel cycle investment positioning — structural long thesis and entry framework
• Risk factors — regulatory uncertainty, geopolitical supply disruption, and timeline slippage
• IU structural investment implications — positioning across the nuclear fuel services value chain