https://github.com/prof-faustus/195-prx
Conditional, falsifiable study of when surface-code distance-suppression collapses (common-mode noise). Theory + GPU simulation.
https://github.com/prof-faustus/195-prx
Last synced: 15 days ago
JSON representation
Conditional, falsifiable study of when surface-code distance-suppression collapses (common-mode noise). Theory + GPU simulation.
- Host: GitHub
- URL: https://github.com/prof-faustus/195-prx
- Owner: prof-faustus
- Created: 2026-06-18T03:22:53.000Z (about 1 month ago)
- Default Branch: main
- Last Pushed: 2026-06-18T07:13:15.000Z (about 1 month ago)
- Last Synced: 2026-06-18T07:25:31.876Z (about 1 month ago)
- Language: Python
- Size: 122 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 195 PRX — Common-Mode Noise & Distance Suppression
Conditional, falsifiable study of when surface-code distance-suppression collapses.
See `PROGRAM.md` for the multi-paper roadmap and `docs/FINDINGS.md` for what is
established. This is theory + simulation; no claim about real hardware is made from
simulation alone.
## Layout
- `src/coherent_sim.py` — GPU-ready coherent-burst state-vector simulator (numpy/cupy).
- `experiments/exp_independent_validation.py` — surface-code threshold sanity (~10%).
- `experiments/exp_rep_correlation.py` — repetition-code: correlation length ξ≳d collapses Λ at fixed error rate.
- `experiments/exp_surface_burst.py` — real surface code: localized burst corrected, spanning burst floors (Paper 1).
- `experiments/exp_coherent_scaling.py` — coherent vs Pauli-twirl LER scaling with d (Paper 2 / Route C).
- `results/` — output CSVs and figures.
## Run (CPU, validation)
```
pip install -r requirements.txt
python experiments/exp_independent_validation.py
python experiments/exp_surface_burst.py
python experiments/exp_coherent_scaling.py # d=3 only on CPU
```
## Run (GPU, production) on the 4-GPU box
```
pip install cupy-cuda12x cuquantum-python quimb opt_einsum
COHSIM_BACKEND=cupy python experiments/exp_coherent_scaling.py # adds d=5 exact
```
d=7 (49 qubits) exceeds exact state vector (2^49). Use the tensor-network backend
(see `docs/GPU_NOTES.md`); the d=3,5 exact numbers are the oracle it must reproduce.
## Validated so far (all reproduced in this repo)
- Surface-code code-capacity threshold ≈ 10% (independent Z, MWPM).
- Fixed-radius burst: b_d → 0, Λ → ∞ (corrected). Spanning burst (ξ∝d): Λ → 1.
- Non-commuting coherent burst at d=3: LER 1.4–2.3× the Pauli twirl (separable decoder).