SMR research

Position in the Physical Layer of AI thesis

SMR is the small-modular-reactor pivot. The Physical Layer of AI map says power is the binding constraint on hyperscaler capacity expansion through 2030. VST is the anchor (existing large-fleet nuclear); SMR is the NEW nuclear capacity pivot. NuScale is the only NRC-design-certified small modular reactor in the US. That regulatory moat is a 5-10 year head start vs every other SMR developer.

Quinn's book opened SMR on 2026-05-05. Snapshot had SMR at +0.68% (just-entered). Sized at 1.6% of book.

Recent catalysts (last 30 to 60 days)

First commercial site selection. UAMPS in Utah formally selected the NuScale 462MWe site for the Carbon Free Power Project. First-of-kind commercial deployment.
DOE Loan Programs Office funding. $1.4B conditional loan commitment for the first US commercial SMR. Federal backstop reduces project finance overhang.
Hyperscaler MOU framework. Disclosed in the Q1 2026 letter: Microsoft and Google both signed exploratory MOUs for SMR capacity reservations on 2030+ delivery.
Q1 print: pre-revenue. SMR is still in the engineering services revenue phase. Real revenue starts when the first commercial unit deploys (~2030 guidance).

The thesis (what has to be true)

1. The Carbon Free Power Project (CFPP) deploys on time. Currently guided to 2030 first power. Slips beyond 2032 break the thesis because alternative SMR designs catch up. 2. NRC retains the NuScale design certification through any administration change. The certification IS the moat. 3. Hyperscaler MOUs convert to firm contracts on 2027 timeframe. The 2030 deployment needs project finance signed by then. 4. NuScale doesn't get acquired at the wrong price. Strategic acquirers (BWXT, Westinghouse, Holtec) all have reasons to bid.

Kill vectors (what would break the thesis)

CFPP cancellation or major slip. The Idaho project IS the proof; if it falls apart, the public-market case collapses.
NRC re-certification process. Any major design change triggers re-review; could add 18-24 months.
Hyperscaler defection to alternative SMRs. X-Energy, Oklo, GE-Hitachi BWRX-300 are all racing for design certification. If one of them catches up faster than NuScale deploys, the head start erodes.
Capital event. SMR is pre-revenue with high capex; a surprise dilutive raise at low price would break the discipline.

Layer context

Power & Grid layer. Sister names: VST (anchor, existing nuclear fleet), BE (distributed fuel cell pivot). SMR is the futures play in the layer; the timeline is longer than VST or BE but the asym is higher if the deployment lands.

Position discipline

Pivot sleeve rules.
Trim trigger: +100 percent from cost.
4-quarter thesis-fail rule applies. If CFPP slips by 2 years or NRC review re-opens, the clock starts.

Moat 7. Asym 8. The regulatory moat is high (NRC certification is a 5+ year process to replicate). The asym is high because the position is pre-revenue with concrete 2030 catalysts.

Real money. Real position. Real receipts.