Skip to content
Technology How it works Breeder — Hyperion Burner — Aegis Burner — MetroVolt AI-Native Architecture Magnets Fuel cycle Safety Roadmap
Solutions AI & Data Centers Defense & Government Grid & Baseload Neutron Detection Quantum
Learn Technical Library
Proof Publications Whitepapers Technical Library Open Science & Reproducibility The Honest Gates
Company About / Mission Leadership Environment Health & Safety Investors Careers Press Contact
3D Model
AI Architecture › L3 · Twin Modeling & AI
L3 · Twin Modeling & AI

L3 for the Breeder: Equilibrium Control

For Hyperion, the central L3 task is holding a valid, stable Grad-Shafranov equilibrium at 9.66 MA in a compact spherical geometry.

THE STACK · click to jumpL7Ecosystem & StrategyL6Experience & VisualizationL5Applications & CopilotsL4OrchestrationL3Twin Modeling & AIL2Data FabricL1Control PlaneL0Foundation▲tlmctl▼L3 · TWIN MODELING & AIThe KRONOS-CTRL digital twin and its predictive shadow.1KRONOS-CTRL Twinlive plant state2GNNscoupled subsystems3PINNsphysics-constrained4Anomaly Ensemblesdrift & fault detection5MPCreceding-horizon control6Predictive Shadowruns seconds aheadMACHINE TIEState estimate descends to L1 control; alerts rise to L4 / L5.KRONOS FUSION ENERGYAI-NATIVE S.M.A.R.T. GENERATORTWIN MODELING & AISHEET 05REV. 2026-08L3 · AI-NATIVE STACK
L3 · Twin Modeling & AI — its place in the stack (left, click any layer) and its internal components (right). Telemetry rises; control descends.

The breeder's defining control problem

The breeder is a compact spherical tokamak, R0 1.2 m, aspect ratio 2.5, 9.66 MA plasma current, 16.84 T peak field. Its performance rests on maintaining a valid MHD equilibrium continuously as pressure and current profiles evolve. Everything else in the breeder twin, neutron source, heat loads, stability margins, is downstream of that equilibrium, so L3's first job is to reconstruct and control it accurately.

How L3 does it

The Grad-Shafranov PINN provides fast equilibrium reconstruction from GNN-conditioned diagnostics; the free-boundary PINN maps coil currents to plasma shape; shape-control MPC drives the poloidal-field and solenoid coils to hold the target boundary; the stability PINN reports margins that constrain the MPC. The predictive shadow projects the equilibrium 50-100 ms ahead so control acts before the plasma drifts.

The spherical geometry sharpens the challenge: strong shaping and high current density mean the equilibrium is sensitive and vertical stability demands fast feedback. Kronos splits timescales, a fast vertical-position loop near L1 for stabilization, a slower multivariable shape MPC above it, so each runs at the cadence its physics requires.

Because the machine is a design/simulation study pre-FOAK, equilibrium control is validated in plant-in-the-loop simulation against reference equilibrium solvers, with the path to synchronize against real reconstruction once FOAK first tritium arrives ~2030. No net-gain behavior is claimed from this control before hardware exists.

Content reviewed August 2026 · design-and-simulation stage