Documentation¶
quchip is an open-source Python toolkit for modelling superconducting quantum chips.
A predictive chip model needs more than a Hamiltonian: device physics, control-line transformations, frames and approximations, dissipation, and measured observables all belong to it. quchip represents each part explicitly. Declare the chip once; the same declaration drives dressed-state analysis, model reduction, control sequencing, open-system simulation, parameter sweeps, and exact JAX gradients.
The declared and resolved physics remain inspectable. chip.unresolved_hamiltonian() shows the authored static model, while chip.hamiltonian() applies the same basis, frame, and approximation strategy used by simulation. A sequence’s Hamiltonian also includes its scheduled drives.
Install¶
quchip requires Python 3.11 or newer.
pip install quchip
Optional extras: quchip[dynamiqs] for the JAX-native backend, quchip[viz] for graph visualization, quchip[scqubits] for scqubits interoperability.
Start with a physical question¶
The guides begin with a small runnable calculation and add one idea at a time:
The Cookbook defines the conventions used by executable quchip examples.
quchip uses GHz for ordinary frequencies, ns for time, and mK for temperature. The implemented conventions and approximations are recorded in the physics reference.
Start from the SQA 2026 talk¶
The post-talk page contains a short runnable snippet and one documentation link for each topic. The guides stand on their own; no knowledge of the presentation is required.
The accompanying paper is quchip: A Differentiable Toolkit for Modeling Quantum Devices (arXiv:2607.17081); citation metadata is in the repository’s CITATION.cff.