SQA 2026 · GOTHENBURG · 26 AUGUST 2026

A differentiable toolkit for modelling quantum devices

Recording, slides, five worked guides, and the paper from the SQA 2026 talk.

slides · PDF the walkthrough on docs.quchip.org
01 · RECORDING

Watch the talk

A differentiable toolkit for modelling quantum devices · 10:58

02 · GUIDES

Five guides from the talk

Start by defining and inspecting a chip. Then move through static spectra, dynamics and readout, chip transformations, and differentiable losses. Each guide starts with a small runnable example and shows its output.

01 · Define and inspect a chip Declare two devices and one coupling, inspect the Hamiltonian, fit dressed targets, and import a scqubits model. 02 · Statics and parameter studies Sweep one parameter, inspect an avoided crossing, then compare fluxonium spectroscopy and readout with published data. 03 · Dynamics, pulses and readout Schedule one pulse, study leakage and open-system dynamics, batch initial states, and read conditional resonator paths. 04 · Chip transformations Rebind and serialize a chip, eliminate modes, partition components, and validate a scheduled active-patch reduction. 05 · Differentiability Differentiate a static loss, fit published fluxonium data, then differentiate one sequence and a shared multi-sequence loss.
03 · READ MORE

The paper and the conventions

What the toolkit implements, how it is validated, and what it deliberately leaves out.

arXiv:2607.17081 quchip: a differentiable toolkit for modelling superconducting quantum chips. PHYSICS.md The implemented conventions and approximations, term by term.

Try it on your own chip

Describe one chip, then analyze, reduce, sequence, simulate, sweep or differentiate it. If quchip cannot yet model your chip, open an issue on GitHub.

$ pip install quchipcopy
source on GitHub
Python ≥ 3.11 · Apache 2.0
Occasional updates on releases and results.