Use cases

Each recipe below is one thing people do with quantcore-circuits, with the steps in short and where to read more. All of them end in the same design.json, so you can combine them.

Draw a qubit by hand in KLayout

For when you want to see and move the geometry yourself.

  1. Install the plugin (plugin) and open KLayout.

  2. Place an Xmon, two bond pads and a readout resonator from the panel; connect them with Connect or Route handles.

  3. Run the Flow strip: Check, LOM, Simulate, Fab (flow strip).

Follow the Q1 + R1 tutorial. Pins and routing: place and route.

Script a chip in Python

For repeatable designs, sweeps and notebooks.

  1. Create a Designer, add components with add, connect them with route.

  2. Call show() to build; with live=True every change is redrawn in KLayout.

  3. Call drc() and lom(); save with save("my_chip.json").

Start with getting started, then the Designer guide and tutorial 0.

Start from target frequencies

For when you know the qubit frequency, anharmonicity, readout frequency and coupling, but not the geometry.

  1. Run qcc synth --f-q 5.1 --alpha -220 --f-r 7.2 --g 90. It prints L_J, the total capacitance, the coupling capacitance and the lambda/4 length.

  2. Turn a circuit file into a layout with qcc circuit layout build/<name>.circuit.json -o build/synth.json.

  3. Open the result in KLayout and refine it.

Limit: this works for one template only, an Xmon with a lambda/4 readout on one feedline. See LOM and tutorial 4.

Check a chip before it goes to fab

For a new or imported design.

  1. Run the DRC: qcc design drc DESIGN (dangling pins, overlaps, crossings, the 100 µm pad rule, die-frame keep-outs).

  2. Run the netlist check: qcc design netlist DESIGN (declared against drawn connections: opens and shorts).

  3. Export the mask: qcc mask DESIGN. The layer map, cheese and corner rounding come from the PDK.

Rules and layers: routing and DRC, PDK, fab export.

Simulate part of a chip and bring the numbers back

For when the lumped estimates are not enough and you have HFSS or Q3D (AEDT 2024 R1 or later).

  1. Define a sim group: the instances and routes to cut out, plus a margin (the plugin’s Simulate dialog, or Designer.sim_group).

  2. Build the job (qcc sim build), solve it (qcc sim run), or render it into an AEDT project without solving (qcc hfss render).

  3. Read the results back (qcc results collect). They fill the LOM, which then uses FEM values.

Details: simulation, results. No converged AEDT run has been made yet; see status.

Tune a design toward its targets

For when the simulated numbers miss the goals you set.

  1. Set targets in the design (for example f_q_GHz and alpha_MHz per qubit, f_GHz and g_MHz per resonator).

  2. Run qcc analyse DESIGN JOB --feedback. The estimators vote, the consensus is compared with the targets, and a new revision is proposed.

  3. Apply it (qcc results apply), rebuild and check again. Open qcc dashboard to follow the residuals.

Details: feedback, dashboard, tutorial 4.

Rebuild a published chip

For learning from a real design or starting from one.

  1. Open the design notebooks. Each one rebuilds a chip from the literature or a design library (Qiskit Metal, KQCircuits, qpdk and others) with the Quantcore library and PDK.

  2. Run its Setup and Design cells; the whole chip appears in KLayout or as a picture.

  3. Change a parameter and rebuild, or continue with the LOM and FEM cells.

How the notebooks are built: notebook template.