ryd-gate

ryd-gate simulates Rydberg neutral-atom systems with continuous-time pulse protocols. The same protocol-bound system can use the exact ODE, MPS, or PEPS backend when that backend supports its level structure and geometry.

The library returns raw physical readouts—observable expectations, final basis amplitudes, and samples. Gate metrics, error budgets, result-analysis plots, optimization, and persistence stay in the calling scripts and notebooks; each concrete protocol's .plot() remains available for inspecting input pulses.

Install

uv pip install -e .                 # exact_ode
uv pip install -e ".[tn]"          # + MPS / DMRG
uv pip install -e ".[tn-2d]"       # + PEPS on CPU
uv pip install -e ".[tn-2d,torch]" # + PEPS on CUDA
uv pip install -e ".[dev]"         # tests and development tools

Quick start

import numpy as np

from ryd_gate import Register, RydbergSystem, level_structure, simulate
from ryd_gate.protocols import SweepProtocol

system = RydbergSystem(
    level_structure=level_structure("1r", ryd_level=70),
    register=Register.square(2, spacing_um=9.0),
    protocol=SweepProtocol(
        t_gate_s=0.5e-6,
        omega_half_rad_s=lambda t: 2 * np.pi * 1e6,
        detuning_rad_s=lambda t: 0.0,
    ),
)

obs = system.observables
n_r = sum(obs.n("r", i) for i in range(system.N))
result = simulate(system, observables={"n_r": n_r})

print(result.times)
print(result.expectation("n_r"))
print(result.sample(shots=1000, seed=0).most_common(3))

The unit and level-label conventions are defined once in Model.

Documentation

• Documentation map: guide, design, ADR, context, and agent-policy placement.
• Model: geometry, level structures, interactions, Hamiltonians, and protocols.
• Simulation: initial states, observables, backends, results, ground states, PEPS evidence, and noise ensembles.
• Gates: CZ/TO/AR and 297-nm workflows, gate metrics, and first-order error budgets.

Public signatures can be inspected with an IDE or help(...); constructors remain the authority for validation and error messages. Runnable examples are in examples; complete research workflows are indexed in scripts/README.md. Project-owned theory and paper drafts are organized under manuscripts/.

Development

OMP_NUM_THREADS=1 uv run pytest -q
uv run ruff check src tests examples
uv run mypy src/ryd_gate
uv run python scripts/check_notebooks.py

References

• Evered et al., “High-fidelity parallel entangling gates on a neutral-atom quantum computer,” Nature 622, 268 (2023).
• Ma et al., “Benchmarking and fidelity response theory of high-fidelity Rydberg entangling gates,” PRX Quantum 6, 010331 (2025).

License

MIT