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

The Mathematics of the Simulation Workloads

A map of the equation families L0 solves, from transport and MHD to transport PDEs, and the numerical methods each demands.

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.

One foundation, several equation families

The L0 workloads are diverse, but they reduce to a handful of mathematical families, each with its own structure and numerical demands. Understanding this map explains why the compute substrate is shaped the way it is: the mathematics of each workload dictates whether it is stochastic or deterministic, explicit or implicit, cheap to parallelize or communication-bound.

The families

Kronos solves: the linear Boltzmann transport equation, stochastically, for neutronics; nonlinear elliptic PDEs, the Grad-Shafranov equilibrium; time-dependent PDE systems, resistive MHD and transport; a high-dimensional kinetic equation, gyrokinetics; and coupled decay-transmutation ODE systems for activation. Each is a distinct mathematical object requiring a distinct method.

Method follows structure

The equation's structure chooses the method. High dimensionality and resonance make transport Monte-Carlo-shaped and perfectly parallel. Ellipticity makes equilibrium a boundary-value problem needing a global solve. Stiffness in MHD forces implicit time integration. Coupling between physics forces the multi-physics strategies. The numerics are not arbitrary; they are consequences of the math.

This map also explains the compute demands. Stochastic and independent workloads weak-scale to huge node counts; implicit coupled workloads are communication-bound and strong-scaling-limited; kinetic workloads are the heaviest of all. The substrate's mix of CPU-dense, low-latency, and GPU-dense partitions is a direct answer to this spread of mathematical shapes.

The same mathematics serves both machines. The breeder and burner differ in geometry and dominant physics, but both are described by these equation families. The equations and methods are shared; the boundary conditions, sources, and parameters make one a spherical tokamak and the other a tandem mirror.

Content reviewed August 2026 · design-and-simulation stage