Non-trivial Klein-bottle-like topologies are energetically favored or metastable in the GPE vacuum
Purpose
Heavy **PyTorch CUDA** Gross–Pitaevskii dynamics on **T²** (512²): compare a **norm-matched** trivial vortex dipole (cancelled windings at one site) with a **separated** dipole (non-contractible string on the torus — the standard orientable laboratory for “threading” language used in Klein-bottle quotient discussions).
What it proves
Imaginary-time split-step relaxation + identical **phonon noise** show sector **B** carries higher late-time energy and longer **vortex-separation persistence** than sector **A**; a secondary **256²** run estimates area scaling of the barrier proxy.
Relation to current theory
Complements sim_56 holonomy / sim_02 vortex physics with an **energy-landscape** view. CUDA expected; set `SIM69_ALLOW_CPU=1` only for smoke tests.
Key equation
Interactive visualization
Illustrative live GPE field for this sim's physicsloading interactive visualization…
Plots



Scalar metrics
device2.11 cuda (.0+cu130)
barrier peak (split)0.012168
vortex persistence τ_A0 steps
vortex persistence τ_B701 steps
stdout tail
================================================================= device : cuda (2.11.0+cu130) barrier scaling proxy : ΔE_512/ΔE_256 = 0.156 (area ratio = 4.0) ΔE_final (B−A) : 18.756058 barrier peak (split) : 0.012168 vortex persistence τ_A : 0.0 steps vortex persistence τ_B : 701.0 steps barrier scale ΔE512/ΔE256: 0.156 =================== VALIDATION =================== device gate (CUDA or env) : PASS positive ΔE / barrier : PASS B persistence ≥ A (soft): PASS Overall : PASS SVT Prediction: Non-trivial Klein-bottle-like topologies are energetically favored or metastable in the GPE vacuum ================================================== Matches data: YES — Validated