What Computing Does for Kronos: An Overview
A single view of how computation runs through every phase of the Kronos program, from first design sweep to plant operation.
One thread through the program
Computing is not one tool used once; it runs through the whole program. It explores the design space, closes the design point, controls the plasma, operates the plant through a digital twin, and assembles the evidence for safety and licensing. The same discipline of reproducibility and verification ties these together.
By phase
- Design: explore the configuration space and close the design point.
- Build: reconcile models with the as-built machine.
- Operate: run control, the digital twin, and maintenance.
- Throughout: safety and licensing evidence and reproducible science.
The flywheel
Each phase feeds the next: design models are corrected by build data and refined by operating data, and those refinements improve future designs. This is the intelligence flywheel, and it is why the AI-native architecture keeps data flowing between phases.
Both machines
The same computing backbone serves the Hyperion breeder, a D-T spherical tokamak, and the Aegis and MetroVolt burner, a D-3He tandem mirror. Their physics differ; the tools, discipline, and data infrastructure are shared.
Honest scope
Both machines are design and simulation efforts today. Construction begins Q2 2027 and first tritium is targeted around 2030, with no hardware net-gain claim before then. Computing is how that design is understood, tested, and made ready, not a substitute for the hardware.
The point
What all this computing does is turn a small team's effort into broad, deep, reproducible understanding of two machines, fast enough to matter and honest enough to trust.