The HALEU Multiplier: Quantitative Dollar-Flow Model from Hyperscaler AI Capex to QLE Revenue
3/9/26
SEQH Capital Research
The HALEU Multiplier: Quantitative Dollar-Flow Model from Hyperscaler AI Capex to QLE Revenue
Tear Sheet – March 7, 2026
Why This Report Exists
The largest capital expenditure cycle in corporate history is underway. The Big Five hyperscalers will spend over 600B dollars on infrastructure in 2026, with approximately 75% (450B) targeting AI-specific infrastructure. This report constructs a layer-by-layer dollar-flow waterfall tracing how each dollar of hyperscaler AI capex converts, through data center power demand, nuclear reactor procurement, fuel cycle economics, and HALEU enrichment, into addressable revenue for Quantum Leap Energy (QLE). The central thesis: Meta is indirectly a QLE customer and doesn’t know it.
The Hyperscaler Nuclear Stampede
Over 10 GW of nuclear capacity has been contracted by hyperscalers in the past 18 months, the largest private-sector nuclear procurement cycle since the original buildout of the 1960s–70s.
Meta: 7.7 GW across four deals, the largest single corporate nuclear buyer in history. TerraPower (2.8 GW / 8 Natrium SMRs / ~17B), Vistra (2.6 GW fleet PPAs), Oklo (1.2 GW Aurora SMRs), Constellation (1.1 GW Clinton plant PPA).
Microsoft: Constellation (835 MW TMI restart / ~16B 20-year PPA).
Amazon: Talen Energy (1.9 GW Susquehanna co-location / ~20B), X-energy (960 MW / 12 Xe-100 SMRs), Kairos Power (500 MW / 7 SMRs).
The HALEU Multiplier Chain
Meta’s 8-reactor TerraPower deal creates deterministic downstream HALEU demand that flows directly to QLE through its definitive 150 MT, 10-year supply agreement with TerraPower.
Natrium fuel requirements:
Initial core load: 15–20 MT HALEU per reactor.
Annual refueling: ~3.6 MT HALEU per reactor.
8-reactor fleet at steady state: 28.8 MT of HALEU annually.
30-year lifetime fuel requirement: approximately 946 MT.
Revenue implications at steady state (all 8 reactors operational by ~2035):
At 15,000 dollars per kg: 432M annual QLE-addressable revenue.
At 25,600 dollars per kg (current NIA benchmark): 737M annually from this single fleet.
Over 30-year plant life, total fuel value exceeds 20B dollars.
ASPI captures 10% perpetual royalty on all QLE revenue: 43–72M annually from Meta deal alone.
QLE’s Cost Moat
QLE’s Quantum Enrichment Process offers a structurally different cost profile than centrifuge alternatives.
Capital efficiency: 100M for a 750,000 SWU facility vs. 4B for equivalent centrifuge capacity, approximately 7.5x lower capital cost per unit of enrichment capacity.
Enrichment selectivity greater than 50, enabling processing of depleted uranium tails that other technologies cannot economically use.
1.7 million metric tons of depleted uranium tails available globally as effectively free (or negative cost) feedstock.
Cost moat widens as primary uranium prices rise, the more expensive traditional feedstock becomes, the more valuable QLE’s tails-processing capability.
The Supply Vacuum
Zero commercial-scale Western HALEU production exists today.
Centrus Energy has produced only 920 kg (less than 1 MT) at its Piketon demonstration cascade through mid-2025.
Russia is the only commercial-scale HALEU supplier globally, but U.S. ban on Russian uranium imports (effective 2028) eliminates this source.
DOE projects cumulative national HALEU needs of 40+ MT by 2030, ramping to 600 MT annually by 2035.
INL models cumulative HALEU demand of 5,350 MT by 2050 under mixed-reactor net-zero scenario.
Key Risks
QLE production timeline execution, enrichment facility under construction but not yet producing at commercial scale.
HALEU pricing uncertainty, market still in its infancy with wide bands.
TerraPower reactor deployment timeline, initial units targeted as early as 2032 but construction risk is real.
Regulatory risk across multiple jurisdictions.
Competitive entry from Centrus expansion, Urenco, or other enrichment providers.
Want the Full HALEU Multiplier Dollar-Flow Model?
[READ THE COMPLETE HYPERSCALER-TO-QLE REVENUE ANALYSIS]
The full report includes a proprietary quantitative dollar-flow model unavailable elsewhere:
Complete hyperscaler capex breakdown: 600B total spend decomposed into GPUs, data center construction, networking, memory, cooling, and power infrastructure layers
Layer-by-layer Sankey flow model mapping each dollar of AI capex through the value chain to QLE-addressable HALEU enrichment revenue
Natrium fuel cycle economics: initial core load, refueling schedule, 30-year lifetime demand profile for the 8-reactor Meta fleet
QLE revenue scenario matrix across multiple HALEU price points (10,000 to 25,600 dollars per kg) with ASPI royalty flow-through
Meta-specific HALEU multiplier analysis showing 4.3x higher conversion ratio vs. broad AI capex average
QLE Quantum Enrichment Process cost moat analysis: 7.5x capital efficiency advantage, tails reprocessing economics, and widening cost moat dynamics
Western HALEU supply vacuum quantification: Centrus 920 kg production vs. 600 MT annual demand by 2035
Demand elasticity model integration: how near-zero price elasticity (ε ≈ 0) amplifies the HALEU multiplier through three reinforcing mechanisms
Goldman Sachs cumulative uranium deficit model (13% through 2035, 32% through 2045) and upstream propagation to HALEU pricing
QLE 150 MT contract valuation at benchmark vs. premium pricing scenarios (2.25B to 3.75B contracted revenue)
Meta committed 17B to TerraPower without knowing it was indirectly buying QLE’s enrichment services. This report traces the dollars and shows you exactly where they land.
