Every visualization below is computed and canon-locked to the ratified freeze (2026-08-27). Drag the controls, watch the machines run. Canon-locked design basis — silent, no financials.
Nested flux surfaces with Shafranov shift. Drag the sliders — the plasma keeps its outward-pointing D-shape (δ < 0), the Hyperion design choice.
Most tokamaks point the plasma's D inward (δ > 0). Hyperion flips it — the D points outward (δ = −0.30). It trades a familiar shape for a calmer, reactor-friendlier edge.
| Conventional +δ | Hyperion −δ | |
|---|---|---|
| Edge instabilities (ELMs) | ELMy | suppressed |
| Divertor heat pulses | transient | steadier |
| Core confinement | good | comparable |
| Reactor robustness | baseline | edge |
Squeeze the aspect ratio and the plasma hugs the center column — a compact, high-pressure shape that does more with less field. Hyperion sits at A ≈ 1.8.
Spherical tokamaks carry high plasma pressure (β) efficiently. The operating point sits safely below the Troyon stability limit — margin, not the edge.
At high pressure the plasma generates a bootstrap current for free — the more it self-drives, the less external power the machine spends holding it up.
Energy confinement improves with plasma current. Hyperion runs at 9.66 MA with an H-mode confinement boost — the CGYRO result pushes it further still.
High-temperature superconductor keeps a huge critical current even at very high field. Drag the temperature — colder means more margin. The conductor peak sits at 26.49 T and still lands well under the critical line.
Every D–T fusion fires a 14.1 MeV neutron into the blanket, where lithium breeds fresh tritium. Bred > burned (TBR 1.42) — the plant makes its own fuel, with a surplus.
Tritium breeding is a design knob. Enriching the lithium-6 in the blanket raises the breeding ratio — the advanced blanket is set for TBR 1.42, comfortably self-sufficient.
Neutron loading isn't uniform — it peaks on the outboard midplane facing the plasma. Mapping it sets the first-wall and blanket design.
All the exhaust power lands on a narrow strip. Flux expansion and detachment spread and cool it — drag the control to bring the peak under the material limit.
Every fusion neutron is a manufacturing tool. In the blanket it makes things the world is short of — starting with tritium, the only bankable base, plus a platform for helium-3 and medical isotopes.
Drag the mirror ratio. The loss cone narrows as the end plugs get stronger — deeper confinement. Hyperion-class burner runs the plugs at Rm = 16×.
Fast ions injected at an angle bounce between the high-field plugs, lingering at their turning points — building pressure where it holds best.
Charged fusion products leave the mirror and fan into the expander. Instead of boiling water, they decelerate against biased grids and give their energy straight up as electricity.
A full shot, start to finish: energize the HTS magnets, break down the plasma, ramp the current, ignite, and hold. Watch the machine come up to power.
The breeder is measured on scientific gain (fusion vs heating); the burner on engineering gain (electricity vs recirculating power). Different yardsticks — both clear the Q = 1 bar.
A staged path: de-risk now, break ground in 2027, first tritium at the start of the next decade, then repeatable burner units toward a fleet.