Hyperscaler Nuclear Demand: Equity Roadmap
3/30/26
SEQH CAPITAL RESEARCH – TEAR SHEET
HYPERSCALER NUCLEAR DEMAND: EQUITY ROADMAP PREVIEW
KEY SETUP
AI data centers are driving a structural power shock, with demand projected at 134–176 GW by 2035 and the nuclear fleet emerging as the primary baseload solution for hyperscalers.
Over ~12 months, Meta, Amazon, Google, and Microsoft have collectively committed to 11–12 GW of nuclear capacity and >$50B of capital across existing fleet PPAs, restarts, and SMR/advanced reactor deals.
SEQH maps ~7.3 GW of near-term contracted capacity from operating reactors and ~20 GW of optioned pipeline, confirming that the hyperscaler nuclear buildout is real, funded, and accelerating.
WHERE THE POWER IS GOING
Meta leads with ~7.7 GW of optioned nuclear (Clinton + Vistra fleet + TerraPower Natrium + Oklo Aurora), followed by Amazon (~3–7 GW via Susquehanna, X‑energy/Cascade, and a 5 GW Xe‑100 consortium), Google (~3.1 GW), and Microsoft (~885 MW).
A granular deal map shows specific assets, technologies, and equity links: CEG (Clinton, TMI), VST (Perry/Davis‑Besse/Beaver Valley, Comanche Peak), TLN (Susquehanna), NEE (Duane Arnold), plus private TerraPower, Kairos, Elementl, CFS, and Helion.
By 2040, SEQH’s “AI Nuclear Demand Stack” sees ~25 GW of nuclear serving data center load, with SMRs providing the bulk of incremental capacity after 2030 and PJM capturing ~60% of nuclear DC capacity.
EQUITY FRAMEWORK – WHO BENEFITS
Tier 1 – Existing fleet owners (CEG, VST, TLN, plus NEE) provide the most immediate, de‑risked upside via long‑dated PPAs and uprates; Vistra and Talen screen as the purest hyperscaler‑contracted nuclear plays.
Tier 2 – Fuel cycle chokepoints (LEU, BWXT, NNE) sit at critical bottlenecks: Centrus is the only U.S. HALEU enricher with a $900M DOE task order and appears in every HALEU‑dependent supply chain; BWXT is the sole U.S. TRISO manufacturer; NNE holds the exclusive INL‑licensed HALEU transport design.
Tier 3 – SMR developers (OKLO, XE pending) offer higher‑beta optionality on the SMR wave, with Oklo’s 1.2 GW Meta deal and 14 GW pipeline highlighted as the most hyperscaler‑integrated profile.
Tier 4 – Uranium producers (UUUU, URG, UEC, NXE, DNN) capture the broader fuel uplift as fleet and SMR demand scale, with names like DNN and NXE flagged for structural volume and cost advantages.
WHERE THE MISPRICING LIVES
LEU is identified as the most mispriced security: down ~63% from its high despite being an irreplaceable HALEU monopoly with DOE, Fluor, and NNSA backing, and effectively gating timelines for TerraPower, Oklo, X‑energy, and Kairos.
TLN and VST are underappreciated hyperscaler nuclear plays, still partially valued as merchant or legacy thermal generators despite multi‑GW, investment‑grade PPAs that reshape their risk/return profiles.
NNE and OKLO are framed as infrastructure and SMR “sleepers” – binary, but with clear databook milestones (INL transport design, Meta/DOE agreements) that sell‑side coverage has yet to fully price.
WHAT PAID MEMBERS GET IN THE FULL REPORT
Upgrade to access the full hyperscaler nuclear roadmap, including:
Complete hyperscaler→asset→equity map and AI Nuclear Demand Stack tables by year and region, plus PJM/MISO/ERCOT/TVA/WECC breakdowns.
Detailed technology stack and fuel‑supply analysis across LEU vs HALEU designs, with explicit dependencies on Centrus, BWXT, and NNE.
The four‑tier SEQH equity basket with company‑level write‑ups, metrics, and rationale, plus side‑by‑side comparisons vs passive nuclear/uranium ETFs.
Mispricing, risk matrix, and 2026–2030 catalyst timeline sections that tie concrete PPA, licensing, HALEU, SMR, and restart milestones directly to specific tickers.
If you want the full deal map, capital‑stack basket, and the mispricing and catalyst work that links hyperscaler PPAs to specific equities, consider upgrading to the paid tier. That unlocks the complete Hyperscaler Nuclear Demand: Equity Roadmap and ongoing updates as AI power demand and nuclear build‑out milestones hit.
