The Infinite Unknown

⚠ Disclaimer: This section may contain incomplete, out of date, or inaccurate entries. It is AI-maintained on a best-effort basis. Do not rely on it as a sole source — verify claims independently using the source materials listed in individual entries.

Overview

Tracks small modular reactor (SMR) developers and advanced nuclear technologies, with particular focus on companies pursuing faster, cheaper deployment models and the emerging AI datacenter power demand that is accelerating commercial nuclear investment. Traditional large nuclear (1 GW+ plants) is covered only where relevant to SMR supply chain or regulatory precedent.

Key Themes

• SMR construction cost and timeline — the core claim every company must prove (target: ~$5,000/kW, 3–5 years; industry history: $13,000+/kW, 10–15 years)
• Datacenter nuclear PPAs as the first commercial demand signal for SMRs
• NRC licensing pathway evolution: Combined License (COL) vs. Construction Permit + Operating License; modular construction resequencing approvals
• HALEU fuel supply chain as a potential chokepoint for non-LWR designs
• Shipyard/modular fabrication as the cost-reduction mechanism — analogous to offshore wind jacket manufacturing
• The 2026 criticality wave and the DOE authorization track. Five reactors reached criticality between June 4 and August 6, 2026 under DOE’s Reactor Pilot Program (EO 14301, May 2025) and the Nuclear Energy Launch Pad: Antares Nuclear (June 4), Valar Atomics (June 18), Deployable Energy (June 30), Aalo Atomics (July 4) and Oklo’s Groves isotope test reactor (Aug 5–6). This is the single largest development in the section since it was created — and the most easily misread. Four of the five were zero-power or near-zero-power test reactors, and none is a licensed commercial plant. DOE authorization is a separate, faster track from NRC commercial licensing and must never be conflated with it: as ANS puts it, “without NRC approval, none of the companies involved will be able to deploy their designs commercially.”
• “Criticality” is a low bar and is being marketed as a high one. Katy Huff (former DOE Assistant Secretary for Nuclear Energy): a zero-power criticality test “can be achieved without making real engineering progress on fuel or design; it simply checks if you can multiply neutrons.” INL Director John Wagner on the Antares milestone: “This is not electricity generation.” Every company entry in this section now carries an explicit note on what its criticality event did and did not demonstrate.
• Capital has decoupled from demonstrated output. Valar Atomics raised $1B at a $6B valuation six weeks after a 100 kWt zero-power test; Antares raised $470M; X-energy IPO'd at ~$9.1B. Meanwhile Deep Fission’s IPO was cut ~75% and trades ~57% below issue, Terrestrial Energy trades ~83% below its post-listing high, and Hadron Energy ~82% below. Private valuations and public marks are pricing the same sector very differently.
• Defence procurement has become a second demand channel alongside datacenters. The Army’s Janus Program (Aug 26, 2026) commits up to $2.2B across five vendors — Radiant (Fort Benning), Antares (Fort Bragg), BWXT (Fort Campbell), General Atomics (Fort Hood), Westinghouse (Fort Drum) — for >20 microreactors, targeting one operating reactor by September 30, 2028. These reactors will be regulated by the Army, contractor-owned and contractor-operated: a third regulatory regime alongside DOE authorization and NRC licensing, with no civilian-nuclear precedent.
• Fuel choice is now the clearest differentiator of schedule risk. Developers using standard LEU and conventional fabrication (Aalo Atomics, Deep Fission, Terrestrial Energy, Blue Energy, Applied Atomics/mPower) sit outside both the HALEU allocation queue and the BWXT TRISO chokepoint. Developers using TRISO or HALEU (Kairos, X-energy, Radiant, NuCube, Antares, Oklo, Valar) do not. Antares’ case is the sharpest: 2028 deployment targets against a Urenco HALEU contract whose UK source does not open until 2031.
• Non-binding pipelines are being reported as capacity. Deep Fission’s own disclosure states its 18.5 GW pipeline is entirely non-binding LOIs “with no commitments to purchase electricity”; Terrestrial’s 4 GW Riot Platforms arrangement is an MOU contemplating a gas bridge; Aalo’s Crusoe partnership and Valar’s Nvidia collaboration are both exploratory. Across the seven newest entries in this section, executed PPAs total zero.
• The public-market cohort arrived in 2026. X-energy (Nasdaq: XE, April), Deep Fission (FISN, June), Hadron Energy (HDRN, May SPAC) joined Terrestrial Energy (IMSR, Oct 2025 SPAC), Oklo (OKLO) and NuScale (SMR). SEC filings now provide a level of financial visibility this sector previously lacked — including going-concern warnings and ineffective disclosure controls at two of them.
• DOE’s Reactor Pilot Program (Executive Order 14301, May 2025) and its successor Nuclear Energy Launch Pad (April 2026) as an accelerant for microreactor demonstration — DOE-authorized criticality/testing at national labs is a faster, separate track from full NRC commercial licensing, and should not be conflated with it
• Regulatory challenge as strategy. Valar Atomics is co-plaintiff with Last Energy and the states of Texas and Utah in a federal suit arguing the NRC lacks licensing authority over reactors below a defence-significance threshold. The Nuclear Innovation Alliance warns a plaintiff win could produce “a patchwork of state-level oversight.”

Deployment Status Tracker

Grid-scale and commercial plants (NRC/CNSC licensed track)

DOE-authorized test reactors — the 2026 criticality cohort

⚑ Read this table carefully. Every entry below reached criticality under DOE authorization, not an NRC licence. Four of the five were zero-power or near-zero-power test reactors that generated little or no electricity. None can sell power commercially.

Other DOE-track and pre-criticality programmes

Companies

Startups & Development Partners

Public Companies

Incumbents

Supply Chain Notes

• Reactor pressure vessels and heavy forgings: Japan Steel Works and Doosan Heavy Industries are primary sources; lead times 3–5+ years; Hitachi relationship gives GEH priority access. New in 2026: TerraPower has added HD Hyundai Heavy Industries as preferred manufacturer for Natrium Reactor Enclosure System components and Hyundai E&C as EPC for up to eight reactors — Korean heavy industry is becoming a significant second source for US advanced reactor fabrication.
• HALEU fuel: Required for Oklo (Aurora), X-energy (Xe-100 via TRISO-X), Kairos (KP-FHR pebbles) and Antares Nuclear (TRISO at 19.75%). Centrus remains the only currently NRC-licensed US HALEU enricher and in July 2026 received a **$900M DOE task order** to expand the American Centrifuge Plant at Piketon, OH — first new commercial capacity expected **2029**, initially targeting 12 MT/yr. X-energy's TRISO-X TX-1 in Oak Ridge is the only US HALEU **fuel fabrication** plant under construction and holds the first NRC Category 2 fuel facility licence to process HALEU. [General Matter](/research/energy/nuclear/general-matter/) holds a $900M DOE award and a 100-acre DOE lease at Paducah but had not received an NRC licence as of last review. Urenco’s UK HALEU facility does not open until 2031.
  • ⚑ The near-term HALEU pool is federally allocated, not commercial. Antares’ Mark-0 ran on DOE-allocated HALEU downblended from NNSA scrap; Oklo’s first Aurora core will use recovered EBR-II material. Multiple 2028-target programmes are drawing against the same finite, politically allocated inventory while their commercial contracts do not deliver until 2029–2031. This is the tightest timing risk in the section.
• TRISO fuel: BWXT (Lynchburg, VA) remains the primary US TRISO particle manufacturer, now supplying Kairos, X-energy, Radiant Industries, NuCube Energy and Antares — five simultaneous first-of-a-kind programmes on one line. BWXT Advanced Fuels LLC (launched Aug 2025) is “exploring” a greenfield TRISO facility; no site, investment figure or timeline has been announced. Two partial alternatives have now emerged: Standard Nuclear (Oak Ridge, TN) fabricated Radiant’s first TRISO delivery in July 2026, and Valar Atomics is building its own co-located TRISO line in Utah on licensed German HOBEG technology.
  • ⚑ BWXT is simultaneously supplier and competitor. It is Antares’ sole fuel fabricator and, since August 2026, a direct competitor to Antares for Army Janus work.
• Standard LEU — the quiet advantage: Aalo Atomics (Urenco USA feedstock, Global Nuclear Fuel fabrication), Deep Fission (Urenco USA, standard PWR assemblies), Terrestrial Energy (SALEU <5%), Blue Energy and Applied Atomics (mPower, <5%) all avoid the HALEU and TRISO queues entirely. Aalo went from fuel contract signature to delivery in roughly a month — a schedule structurally impossible on either constrained pathway. ⚑ Hadron Energy is the exception that proves the rule: its 8% “LEU+” sits above the standard 5% ceiling, and no US commercial LEU+ enrichment or fabrication line operates at scale.
• Fuel salt (new layer): Terrestrial Energy’s IMSR requires uranium dissolved in fluoride salt — a first-of-a-kind industrial process with no commercial precedent. Being addressed via Westinghouse at Springfields, UK and DOE’s Project TEFLA fuel-line pilot. Neither facility exists yet.
• Li-7 enrichment: Required for Kairos’s fluoride salt coolant (FLiBe); no US commercial Li-7 enrichment capability currently; DOE/ORNL developing domestic capability.
• Nuclear-grade graphite: Moderator for Aalo, Antares, Valar, Terrestrial and the HTGR designs generally. X-energy is investing up to $8M with SGL Carbon to double EU NBG-18 billet capacity by 2030 (enough for up to 8 reactors/yr); Aalo sources from Amsted Graphite; NRG Petten conducts irradiation testing for Terrestrial. Graphite behaviour under sustained flux is also a recurring open technical question for these designs.
• Turbine generators: GE Vernova, Siemens Energy (X-energy/Oklo named); Baker Hughes supplying Aalo’s 10 MWe set and engaged by Terra Innovatum on sCO2 conversion; standard power gen supply chain, less constrained than nuclear-specific components.
• Instrumentation & control — an emerging shared dependency: Paragon Energy Solutions (a Mirion company) now appears across three unrelated programmes in this section — NuScale (HIPS platform: Module Protection System, Safety Display and Indication System, Plant Protection System, contracted June 2026), Aalo Atomics, and Hadron Energy. Worth tracking as a potential single-vendor concentration at the I&C layer.
• EPC and construction: Kiewit appears twice (Oklo’s Aurora at INL, Valar’s Utah site). Bechtel (TerraPower), Hyundai E&C (TerraPower, up to 8 reactors with completion/price/performance guarantees), Day & Zimmermann (Deep Fission, above-ground scope only), Clark Construction (TRISO-X TX-1), Fluor and Geiger Brothers (Centrus Piketon expansion). ⚑ Deep Fission has no named drilling contractor despite drilling being its entire critical path.
• Sodium systems expertise: TerraPower (Natrium), Oklo (Aurora), Aalo Atomics (thermal-spectrum pool) and Antares (heat pipes) all use sodium in some form; limited number of qualified fabricators and constructors with sodium reactor experience in the US.
• Shared federal test infrastructure: DOE/NRIC’s DOME test bed and the Materials and Fuels Complex at Idaho National Laboratory are shared by multiple companies under the Reactor Pilot Program and Nuclear Energy Launch Pad — Antares, Deployable Energy, Aalo Atomics and Radiant Industries have all tested or are testing at INL. DOE authorization to test there is a separate, faster pathway than NRC commercial licensing and should not be read as commercial approval.

Subsections

Individual company entries are listed in the tables above. See the Energy section for adjacent research on grid markets, batteries, solar and virtual power plants.