QPDK is an open-source process design kit (PDK) for superconducting quantum RF applications built on gdsfactory. It provides a library of parametric quantum circuit components (transmon qubits, CPW resonators, Josephson junctions, etc.), analytical S-parameter models, routing utilities, and test-chip examples.
QPDK gives researchers, engineers, and students a scriptable, version-controlled foundation to go from concept to GDSII in minutes.
- Rich component library — Transmons, fluxonium, unimon and gatemon qubits, CPW resonators, interdigital capacitors, SQUID junctions, launchers, bump bonds, TSVs, and more.
- Parametric & composable — Combine Python functions (
@gf.cell) into hierarchical designs or define full chips in YAML. - Analytical circuit models — Fast, differentiable S-parameter simulations powered by SAX and JAX.
- Automated routing — CPW-aware routing strategies with auto-tapers for complex layouts.
- KLayout integration — Layer definitions, technology files, and cross sections for immediate visual inspection.
- Regression-tested — GDS regression tests, netlist checks, and model validation.
- Notebook-driven workflows — Jupyter notebooks for frequency modeling, tolerance analysis, parameter extraction, pulse-level simulation, and optimization.
- GDSFactory+ integration — Seamlessly design, verify, and validate chips with the enhanced commercial extension. Includes access to 43+ foundry PDKs, graphical layout and schematic editors directly in VSCode, an AI assistant, and comprehensive verification tools (DRC, LVS, Connectivity checks). See the QPDK graphical workflow tutorial.
QPDK ships with a broad set of ready-to-use superconducting circuit components. Browse all the components in the documentation.
| Transmon | Fluxonium |
|---|---|
| Double-pad capacitively shunted qubit | Superinductance-shunted qubit |
 |
 |
| Unimon | SQUID Junction |
|---|---|
| Resonator-embedded junction qubit | Superconducting quantum interference device |
 |
 |
| CPW Resonator | Interdigital Capacitor |
|---|---|
| Meandering coplanar waveguide resonator | Finger-style lumped-element capacitor |
 |
 |
Transmon with Resonator & Probeline — Qubit cell with coupled resonator and probeline section:
QPDK includes complete, tapeout-ready test chip examples that demonstrate real-world design workflows.
A four-transmon test chip with coupled readout resonators, probeline routing, flux lines, and launchers. Defined entirely in YAML.
The same qubit test chip with magnetic vortex trapping holes filling the ground plane and chip edges.
A four-flipmon test chip demonstrating flip-chip qubit integration with coupled readout resonators, interdigital capacitor inputs, probeline routing, and flux lines — shown here with ground-plane filling and chip edges.
A two-gatemon test chip after Riechert et al. (2025): each gatemon sits at the input end of a half-wave meander resonator between two series coupling capacitors, with its gate electrode routed to a dedicated launcher. Shown here with ground-plane filling and chip edges. Both independent readout channels target the reported device A frequency of 5.056 GHz using QPDK's analytical CPW and capacitor models, including the nominal island load with its Josephson junction open. The physical lengths and capacitor dimensions are design estimates, and the qubit coupling still needs electromagnetic calibration.
Two mirrored CPW lines that use every metal layer of a flip-chip stack. Each line runs from a launcher on M1 through a TSV down to a backside (MB) underpass, back up to M1, across an indium-bump transition to the top chip (M2), through a top-chip TSV to a meander on the top chip's backside (MB2), and back out the same way to a second launcher. Indium bump arrays tie the two chips' ground planes together.
A 3D cut-away through one of the lines, rendered with to_3d() and translucent substrates, shows the TSVs through both
chips and the top chip with its MB2 meander on the backside.
A 16-resonator characterization chip with systematically varied CPW widths and gaps across two probelines. Ideal for extracting loss tangents and kinetic inductance. See also the notebook demonstrating network model simulations and fabrication tolerance Monte Carlo.
QPDK comes with worked examples for circuit, quantum, and electromagnetic simulation. Every figure below is taken from one of the example notebooks as rendered in the documentation. Click a title to open the full notebook.
| Resonator Test Chip Transmission | Fabrication Tolerance Monte Carlo |
|---|---|
| Both probelines of the 16-resonator chip, via SAX | Resonance spread from CPW width and gap variations |
 |
 |
| Flux-Tunable Transmon Spectrum | Pulse-Level Gate Simulation |
|---|---|
| Energy levels versus external flux with scQubits | X-gate populations and leakage in QuTiP-QIP |
 |
 |
| Dispersive Shift | Differentiable Transmon Optimization |
|---|---|
| Block-diagonalized dispersive shift with Pymablock | Circulax JAX gradients to target f and anharmonicity |
 |
 |
| Traveling-Wave Parametric Amplifier | Flux-Pumped JPA |
|---|---|
| 2047-junction TWPA gain in JosephsonCircuits.jl | VACASK harmonic balance vs. single-mode theory |
 |
 |
| COMSOL Resonator | HFSS Eigenmode | Elmer Capacitance |
|---|---|---|
| Driven E-field at resonance | Meandered CPW eigenmode field | IDC capacitance vs. FEM remeshing |
 |
 |
 |
COMSOL Transmon Capacitance — Electrostatic potential and field around the double-pad transmon:
The same workflows extend to HFSS driven-modal and Q2D impedance analyses, Qucs-S model validation, and NetKet and Optuna optimization.
import gdsfactory as gf
from qpdk import PDK
PDK.activate()
# Create a transmon qubit
from qpdk.cells.transmon import double_pad_transmon
qubit = double_pad_transmon(pad_size=(250, 400), pad_gap=15)
qubit.plot()# Build a complete test chip from YAML
from gdsfactory.read import from_yaml
from qpdk import tech
chip = from_yaml(
"qpdk/samples/qubit_test_chip.pic.yml",
routing_strategies=tech.routing_strategies,
)
chip.show() # Opens in KLayoutPrefer a visual workflow? QPDK runs in GDSFactory+, where you can place components in the schematic editor, inspect the layout, and run circuit simulations without leaving VSCode. The Quantum-RF graphical workflow tutorial walks through exploring and simulating a resonator with QPDK step by step.
- PDK cells in the documentation — Collection of all available geometries.
qpdk/samples/— Example layouts and simulations including qubit test chips, resonator arrays, routing demos, and 3D export.notebooks/— Jupyter notebooks covering:- Resonator frequency modeling and S-parameter analysis
- Circuit simulation with SAX
- Monte Carlo fabrication tolerance analysis
- Hamiltonian parameter extraction with scqubits
- Pulse-level quantum gate simulation with QuTiP
- Capacitor geometry optimization with Optuna
- Dispersive shift calculation with Pymablock
- Transmon design optimization with NetKet
- GDSFactory+ Quantum-RF tutorial — Graphical design and simulation of a resonator with QPDK in GDSFactory+.
- gsim example notebooks — Electromagnetic simulation examples using Palace and Meep with gdsfactory.
We recommend uv for package management and use just for
project-specific recipes.
Install the package with:
uv pip install qpdkOr with pip:
pip install qpdkInstall the optional dependencies for the analytical models and simulation tools (SAX, scqubits, JAX) with:
uv pip install "qpdk[models]"models is one of several extras — hfss, circulax, netket, pymablock, qutip, ray, scqubits and graphics
pull in the remaining simulation backends, and extras compose ("qpdk[models,netket]"). See
Installing optional extras for
what each one installs and which notebooks need it.
To use the PDK in KLayout (for viewing GDS files with correct layers and technology settings), install the technology files:
python -m qpdk.install_techNote
After installation, restart KLayout so the new technology appears.
Follow the installation and development workflow instructions.
quantum-rf-pdk/
├── qpdk/ Core Python package
│ ├── cells/ Component definitions (transmons, resonators, capacitors, …)
│ ├── models/ Analytical models, mostly S-parameters
│ ├── samples/ Example layouts and complete test chips
│ ├── klayout/ KLayout technology files
│ └── tech.py Layer stack, cross sections, routing strategies
├── tests/ Regression, integration and unit tests
├── notebooks/ Jupyter notebooks for design and simulation workflows
└── docs/ Sphinx documentation (HTML + PDF)
If you use QPDK in research, please cite the Zenodo record for the release you used. Use the concept DOI when referring to QPDK as a project without a specific version. Citation metadata is in CITATION.cff. See the authorship policy for how citation authors are selected.
We welcome contributions of all sizes: new components, improved models, bug fixes, documentation, and notebook tutorials. Please see the contributing guide to get started.
To build a PDK for your own fabrication process on top of QPDK, see Customizing QPDK for Your Process.
For commercial support, training, and custom PDK development, please visit gdsfactory+ (gdsfactory.com/plus).
QPDK is released under the MIT License.