From Simulation to Hardware

The breeder's design is a body of simulation that the built machine will validate gate by gate, measured against pre-registered predictions.

What validation means here

Today the breeder exists as simulation: plasma physics models, magnet and structural analysis, and neutronics for breeding. Validation is the process of comparing what these models predict to what the built machine actually does. The gate program is the structure of that comparison.

The validation chain

• Confinement models validated at G1 against measured energy confinement
• Magnet and structural models validated at G2 against measured field and margin
• Neutronics and breeding models validated at G3 against measured yield and TBR
• Integrated behavior validated through commissioning

Why simulation is not enough

Simulation bounds behavior but cannot settle it, especially for confinement scaling and three-dimensional net breeding, where the gap between model and reality is a known risk. Only the hardware settles these. This is why the program does not present simulation results as achievements, and why it builds a first-of-a-kind unit at all.

Reporting discipline

Validation is reported measured-versus-predicted, using the pre-registered predictions. A model that predicted the machine well is validated; one that missed is corrected and feeds the NOAK design. Either outcome advances the program, which is why the discipline matters more than any single number matching.

All current figures are simulation-based design values for a machine not yet built. Validation converts them, one gate at a time, into demonstrated results.