Hyperion as a Strategic-Materials Platform: Gain, Tritium Breeding, He-3 Co-Production, and 14 MeV Neutron Output from a Compact Spherical-Tokamak

Fusion breeders have historically foundered on two coupled obstacles: net electricity forces a large plasma and a large blanket, and a self-sufficient closed fuel cycle at power-plant scale implies a dangerous standing tritium inventory. Both are consequences of demanding electricity from a device whose neutrons are physically its most valuable output. We present the integrated physics of Hyperion, a compact negative-triangularity spherical-tokamak breeder sized to a strategic-materials mission — tritium, helium-3, and 14 MeV neutrons — and show that once the net-electricity requirement is dropped, both obstacles fall below their binding thresholds. The frozen design point (I_p=9.66 MA, on-axis B_0=8.0 T, centrepost peak field 16.84 T, δ=-0.30, κ=2.0, R_0=1.20 m, A=2.5) closes its zero-dimensional power balance at a plasma gain Q=3.076 and fusion power P_fus=85.04 MW, carried openly as a driven, sub-ignition design-point target reached at demonstrated confinement. We state the governing equations of every subsystem as explicit mathematics — the Bosch–Hale power balance, the free-boundary Grad–Shafranov equilibrium and its ideal-MHD margins, the Sauter neoclassical bootstrap, the nonlinear-gyrokinetic transport ratio, the neutron-transport tritium-breeding functional, the tritium and helium-3 mass-balance ordinary differential equations, and the Rosenbluth–Putvinski runaway avalanche — carry each derivation from its first principle to its discretized numerical form, and evaluate them on a verification-first computational stack (Kronos toolkit, FreeGS/FreeGSNKE, NEO, nonlinear CGYRO, TGYRO, ASCOT5, OpenMC on ENDF/B-VIII.0, and a NIMROD-verified disruption model), with the nonlinear gyrokinetics and Monte-Carlo neutronics executed on the NVIDIA GPU HPC campaign. Negative triangularity is the confinement lever: converged nonlinear gyrokinetics show it suppressing ion-scale transport by 40–45% and electron-scale (ETG) transport by 79–80% relative to a positive-triangularity twin, with the operating point subcritical and the advantage robust (56–66%) across a factor-sixteen collisionality scan. The free-boundary equilibrium is Mercier-, ballooning- and external-kink-stable with a 1.64× normalized-beta margin to the no-wall Troyon limit; the current is 85% externally driven with a 15.5% bootstrap fraction; fusion-born alphas are 99.98% confined; and tritium breeding rises along a confirmed ladder from a net surplus of 0.29 to ~2.0 kg-T yr^-1 per unit at the advanced blanket (net breeding ratio 1.42), making a single advanced unit a tritium exporter and a national-scale 4 kg-T yr^-1 supply an explicit fleet property. The inboard centrepost is a quantified consumable cartridge, and the disruption response is answered directly and is demanding: a severe runaway-electron avalanche makes shattered-pellet mitigation a hard design requirement. We close with a dedicated sensitivity/uncertainty section and an experiment-and-theory benchmarking section. Every headline number is a frozen anchor with a gate serial, independently re-run and stamped in the accompanying physics de-risking register. Key results (frozen anchors): I_p = 9.66 MA; P_fus = 85.04 MW; Lambda_T = 4.77 kg-T; E_DT = 17.59 MeV; E_α = 3.52 MeV; E_n = 14.06 MeV. Methods & codes: Kronos toolkit, FreeGSNKE, FreeGS, NEO, CGYRO, TGYRO, TGLF, ASCOT5, OpenMC, NIMROD, SOLPS-ITER, SOLPS, DAGMC, ENDF/B-VIII.0, Bosch–Hale, Sauter, Rosenbluth–Putvinski, STEP. Live verification: 32 gate validator(s) with live-recompute cards (30 reproduced, 2 estimated). See the Verification section and data/verification.csv. Related: Paper page · De-risking register · 3D model · Learn more about Kronos Part of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689). All numerical values are frozen design-point anchors; see the register. Public research artifact. No proprietary, financial, or supply-chain information is included. Previous version: DOI 10.5281/zenodo.22132235. Please cite this version.