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 › L0 · Foundation
L0 · Foundation

Tritium Breeding Ratio Parameter Sweeps

Wide, independent Monte Carlo campaigns that map breeder blanket design against the TBR lever set 1.1, 1.5, and 1.8.

THE STACK · click to jumpL7Ecosystem & StrategyL6Experience & VisualizationL5Applications & CopilotsL4OrchestrationL3Twin Modeling & AIL2Data FabricL1Control PlaneL0Foundation▲tlmctl▼L0 · FOUNDATIONThe offline compute substrate — multi-physics & batch training.1Cloud HPCelastic burst2Bare-Metal ClusterGPU / CPU3Supercomputingmulti-physics runs4Batch Trainingmodel builds5Simulation FarmGrad-Shafranov · MHD6Object StorecheckpointsMACHINE TIETrains the models that ship UP to L3 — no real-time path to the machine.KRONOS FUSION ENERGYAI-NATIVE S.M.A.R.T. GENERATORFOUNDATIONSHEET 02REV. 2026-08L0 · AI-NATIVE STACK
L0 · Foundation — its place in the stack (left, click any layer) and its internal components (right). Telemetry rises; control descends.

TBR as a swept design lever

Kronos treats the breeder's tritium breeding ratio as a design lever rather than a single fixed target, studying it across 1.1, 1.5, and 1.8. A parameter sweep on L0 launches many independent neutronics runs, each with different blanket composition, lithium-6 enrichment, multiplier fraction, and structural layout, and scores the TBR each configuration achieves.

An embarrassingly parallel campaign

Each configuration is independent, so the sweep is embarrassingly parallel: thousands of Monte Carlo jobs run concurrently with no cross-talk, ideal for elastic cloud HPC. The campaign fans out, each job traces its histories to a variance target, and results are gathered into a response surface over the design space.

Reading the response surface

The output is a map from blanket design to achievable TBR, with statistical error bars from the Monte Carlo. Configurations that reach 1.5 or 1.8 cost geometry and material choices that the sweep makes explicit, so the design team trades breeding margin against practicality with quantified confidence rather than intuition.

Interesting regions of the surface are then re-run on certified nodes under strict reproducibility before any value is trusted for design. Cloud explores the space broadly; bare-metal certifies the specific points that matter. A TBR figure that will drive design must be replayable bit-for-bit.

The sweep result is more than a number. It quantifies how sensitive breeding is to each lever, which tells the fuel-cycle model how much margin exists and tells the twin's isotope-balancing surrogate how breeding responds to configuration. The lever set 1.1, 1.5, 1.8 is a deliberate span from marginal to comfortable, chosen to bracket the honest engineering range.

Content reviewed August 2026 · design-and-simulation stage