Noetic Machine

Computational physics toolbox for SU(2) gauge models and topological defect structures.
Calibrated on the Bogomolny–Prasad–Sommerfield bound; predictive over three orders of magnitude.

---

What this is

A reproducible computational environment for testing field-theoretic structures. Given a candidate model (ansatz, Lagrangian, symmetry-breaking pattern) and measured anchoring data (physical constants, scales), it returns a verdict — existence, spectrum, quantisation, regime — with documented invariants, frozen protocols, and SHA-256 hashed artefacts.

This is not a simulator. It does not answer "what happens if…" (trajectory). It answers "what exists, where, at what cost" (structure, regime, constitutive law).

Five confirmed predictions (Georgi–Glashow SU(2) banc)

#	Prediction	Key result	Status
P0	BPS mass calibration	C(ρ=1) = 1.3098 (lit. 1.24–1.31); C(ρ=0.5) = 0.9981	✅ Confirmed
P1	Bound-state spectrum (Bohr atom)	Coulomb pure to 10⁻⁴, LRL degeneracy, a₀ = 137 l.u.	✅ Confirmed
P2	Dirac charge quantisation	e·g = 2π exact, n = 1, ratio 1.000000	✅ Confirmed
P3	Nucleus–ring coexistence + flux tube	ΦM = 4π = Φv(Q=2), confinement emergent	✅ Confirmed
P4	Phase diagram (g, v, ρ)	2 regimes, boundary ρ* ≈ 0.75, auto-dual point	✅ Confirmed

Related Repositories

Repository	Role
noetic-applications	14 experimental case studies (P7–P20) applying the finite-core solver to atomic, nuclear, particle, condensed-matter, and molecular physics
spectral-triple-minimality	Mathematical foundations — 4 theorems (dimension, k-bound, margin-3, non-uniqueness) and the KO-6 arithmetic law
ko6-spectral-solver	Spectral benchmarks B1–B3 (Taylor–Green, KdV, Ising 2D)

Citation

@software{noetic_machine,
  author = {Portemann, Patrice},
  title = {Noetic Machine: A Non-Perturbative Finite-Core Solver},
  url = {https://github.com/PORTEMANN/noetic-machine},
  version = {1.0},
  year = {2027}
}

See CITATION.bib for cross-repository entries.

---

Repository structure

.
├── README.md                 # This file
├── LICENSE                   # MIT
├── MANUAL.md                 # Engineer's manual (full protocol)
├── CITATION.bib              # Cross-repo BibTeX entries
├── src/
│   ├── core/                 # Numerical cores (vortex, entanglement)
│   ├── benchmarks/           # B1–B5: rotation, soliton, Landau, turbulence, oscillator
│   ├── modules/              # D1–D4: mediator, rings, dispersion, vortex pairs
│   │                           E44–E48: nucleation, capture, conservation, assembly
│   └── off_corpus/           # P0–P4: BPS calibration, spectrum, Dirac, bridge, phases
├── protocols/                # Frozen JSON protocols (50+ benchmark and module configs)
├── data/                     # Run outputs (.pkl, .json, .csv) — generated locally
├── notebooks/                # Verification notebooks
└── docs/
    ├── manual/               # Engineer's manual (full)
    └── results/              # P0–P4 result notes (markdown)

Quick start

Requirements

• Python ≥ 3.10
• NumPy, SciPy, Matplotlib
• (optional) Jupyter for notebooks

Run a benchmark

python src/benchmarks/b1_analyse.py

Run off-corpus P0 (BPS calibration)

python src/off_corpus/p0_monopole_su2.py

Verify gradient correctness

Every module must validate its discrete functional gradient against finite differences before production:

assert np.allclose(grad_discrete, dx * grad_continuous, rtol=1e-2)

Protocol discipline

1. Frozen protocol — ansatz, grid conventions, tolerances, and verdict criteria are fixed before execution. No post-hoc adjustment.
2. Execution — variational minimisation or exact diagonalisation, with analytically exact discrete gradient.
3. Verdict — read from invariants (dimensionless mass, core radius, virial decomposition, charge product, flux). Never from raw curves.
4. Fingerprints — every artefact (script, data, figure) is SHA-256 hashed; verdict notes cite prior fingerprints.
5. Published failures — negative verdicts and numerical artefacts are published with the same care as successes.

Parameter classification

Class	Meaning	Examples
Derived	Output of the machine	C(ρ), spectra, radii, e·g = 2π
Measured	Physico-chemical anchoring	α = 1/137.036, R_core = 3.04 l.u.
Assumed constitutive	Structural choice, ontology	ρ = λ/e², Georgi–Glashow model itself

Author

Patrice Portemann

A result that does not reproduce on another machine does not exist.