Gold = breakout month. Green = months above structural threshold (≥25/100). Orange = below threshold. Source: Google Trends CSV, downloaded August 12, 2026. August 2026 reading from Zorthex dataset revision.
t_start selected: May 2005
Source: IAEA TECDOC-1451 — "Innovative Small and Medium Sized Reactors: Design Features, Safety Approaches and R&D Trends" · Level A (international institutional publication)
Rationale: First institutional document formalising SMR as a reactor category with defined parameters (≤300 MWe, modular design, factory-fabricated). This moment marks when SMR transitioned from individual research projects to a recognised institutional category with international standards.
DOE Nuclear Energy Research Initiative funds early SMR concepts. Oregon State University research (basis for NuScale). Conceptual phase — no formalised category exists.
IAEA TECDOC-1451 published. First international institutional framework defining SMR as a category. This is t_start. Level A.
NuScale Power founded. First company exclusively focused on SMR commercialisation. Level B.
First Google Trends signal: 3/100. The term enters measurable search behaviour.
DOE SMR Licensing Technical Support program announced. $452M over five years. Level A.
NuScale receives NRC Standard Design Approval — the only NRC-certified SMR design. Level A.
Amazon anchors $500M investment in X-energy. Google, Microsoft announce nuclear PPAs. Google Trends rises to 13/100.
TerraPower closes $650M Series C (Bill Gates, NVIDIA NVentures). X-energy raises $700M. $1.3B total equity in 2025.
Public attention breakout. Google Trends: 16 → 100 in one month. t_peak confirmed.
X-energy IPO: $1.02 billion. First pure-play SMR IPO. Holtec files S-1 targeting $10B+ valuation.
Deployable Energy Unity reaches criticality — third microreactor under DOE July 4 directive. Bank of America estimates $10 trillion nuclear opportunity. Last month above threshold (44/100).
Score drops to 23/100 — below threshold. First month below since breakout. 12 consecutive months completed exactly. STRUCTURAL classification locked. Operational status: Receding.
Three-source convergence confirmed. All three sources show directional alignment from August 2025. Google Trends documents the breakout and the August 2026 retreat. Wikipedia documents sustained informational demand. Reddit documents the transition from technical to financial mainstream discourse. Family classification: Intermediate (GT high then receding, Wiki moderate/event-driven, Reddit matura).
Google Trends geographic data, 2004–2026. National-level readings.
Geographic reading. South Korea dominates (100) — driven by KAERI's SMART SMR programme. Canada (40) reflects the Darlington BWRX-300 project, the first SMR under construction in North America. The US (20) is surprisingly low relative to its investment volume — likely because American search behaviour is distributed across company names (NuScale, Oklo, TerraPower) rather than the generic category term.
For 20 years, the IAEA published technical standards, NuScale obtained the only NRC design certification, TerraPower secured $650M from Bill Gates and NVIDIA, X-energy raised $700M from Amazon, and China built the world's first commercial land-based SMR. All of this happened while Google Trends showed 2–7 out of 100.
The breakout in August 2025 closed that window. It was triggered by a convergence of signals: AI data centre energy demand, DOE executive orders mandating three operational reactors by July 4, 2026, and $1.3 billion in equity investment in a single year.
What the Receding status adds: The window opened and closed in exactly 12 months. The public discovered SMR, processed it, and moved on. The attention lasted long enough to qualify as structural (the term entered the vocabulary) but not long enough to sustain active interest. This is the signature of a phenomenon that was catapulted into public awareness by a specific narrative (AI energy demand) rather than by sustained consumer engagement. When the narrative moves on, the attention follows.
For comparison: RWA Tokenization (STRUCTURAL · Active, current 40) and Stablecoins (STRUCTURAL · Active, current ~31) are both above threshold. Their attention is sustained because the underlying activity generates continuous search demand. SMR's attention was driven by investment narrative, not operational demand — nobody Googles "small modular reactor" because they're buying one. The Receding status documents this structural difference.
| Phenomenon | L (months) | Classification | Op. Status |
|---|---|---|---|
| Digital Nomad | 325 mo | STRUCTURAL | Active |
| Small Modular Reactor ◀ | 243 mo | STRUCTURAL | Receding |
| Zero Day Vulnerability | 231 mo | STRUCTURAL | Active |
| Streaming | 218+ mo | STRUCTURAL | Dominant |
| iPhone | 208 mo | STRUCTURAL | Dominant |
| Green Hydrogen | 204 mo | STRUCTURAL | Active |
| Super App | 185 mo | STRUCTURAL | Active |
| Autonomous Vehicle | 176 mo | STRUCTURAL | Active |
| Cybersecurity | 145 mo | STRUCTURAL | Active |
| Stablecoins | 128 mo | STRUCTURAL | Active |
At 243 months, SMR holds the second-longest lag in the dataset and the longest in the Energy domain. The only case with a longer lag (Digital Nomad, L=325) is Active. SMR is Receding — making it the longest-lagging Receding case in the dataset. The regime is a hybrid: Institutional Mass + Policy-Trigger, which is rare (most cases show one mechanism, not both).
The original report set this condition: "IF attention drops below 25 → Monitor for regime shift to Receding." Triggered. Google Trends dropped to 23/100 in August 2026 — the first month below threshold since the breakout. Operational status moved to Receding. The next question is whether this is a dip (the score returns above 25 in September–October) or a permanent decline. If below 25 for 3+ consecutive months, the Receding status is confirmed as sustained.
X-energy IPO ($1.02B, April 2026) completed. Holtec filed S-1 targeting $10B+ valuation. Each SMR IPO functions as a public-attention amplifier. A Holtec listing could trigger a secondary attention spike. Monitor NuScale earnings trajectory (NYSE: SMR) for execution signal.
No Western SMR has yet delivered commercial electricity. China's ACP100 status unclear. Deployable Energy Unity reached criticality (July 2026). If a Western SMR delivers commercial power, expect a fundamental attention reset — the narrative shifts from "promise" to "operating asset."
NuScale's UAMPS cancellation (2023) is the historical precedent: project failures produce temporary attention spikes followed by sentiment inversion. A major SMR project cancellation or cost overrun (>$1B) during the Receding phase could accelerate the attention decline permanently.
| Data | Source | Level | Date |
|---|---|---|---|
| t_start definition | IAEA TECDOC-1451 | A | May 2005 |
| Google Trends CSV | trends.google.com · "small modular reactor" · Worldwide | Primary | Aug 12, 2026 |
| Wikipedia pageviews | Wikimedia API · daily | Primary | Aug 2024 – Aug 2026 |
| August 2026 verification | Zorthex dataset revision 71/71 | Primary | August 2026 |
| NuScale NRC certification | NRC Standard Design Approval | A | Jan 2023 |
| 2025 equity investment ($1.3B) | Industry aggregation | C | May 2026 |
| X-energy IPO | VanEck investment analysis | B | Apr 2026 |
| Bank of America $10T estimate | Motley Fool / Yahoo Finance (BofA research note) | C | Jul 2026 |
Data availability: Google Trends CSV available at github.com/zorthex2026/zorthex-diffusion-lag · Methodology: zorthex.com/methodology · Contact: zorthex.official@gmail.com