# 12-qubit flux processor (Litho) Twelve flux-tunable Xmons in two columns of six, one each side of a vertical feedline. Each has a λ/4 readout (6.05 to 6.82 GHz in 70 MHz steps) and a flux line that runs out to a side pad. The layout is from Qiskit-Metal-to-Litho (CC BY 4.0), rebuilt on the Quantcore PDK with a trilayer SQUID per qubit. This notebook is an example of a larger chip: twelve readouts and twelve flux lines, with the LOM reading the flux lines as bias lines. ```{image} ../../images/designs/community/qm2litho_12q_flux_processor.png :alt: The whole qm2litho_12q_flux_processor chip as the v2 notebook builds it :width: 100% ``` ## At a glance | | | |---|---| | Notebook | `examples/designs/community/qm2litho_12q_flux_processor.ipynb` (an older one-cell version is kept as `qm2litho_12q_flux_processor.v1.ipynb`) | | Source | Qiskit-Metal-to-Litho `12_Qubit_Flux_Tunable_Processor_in_Colab.ipynb`, CC BY 4.0 | | Chip | 11 × 11 mm Quantcore die (the source's chip size), feedline 11.7 / 5.1 µm | | On the chip | 14 pads · 1 feedline · 12 flux-tunable Xmons · 12 λ/4 hanger readouts · 12 flux lines | | Builds today | yes (layout built on 2026-11-19 in 10 s) | ## How the notebook builds it ```{image} ../../images/designs/community/qm2litho_12q_flux_processor_steps.png :alt: The layout segments of qm2litho_12q_flux_processor, one picture each :width: 100% ``` 1. **Die.** The 11 × 11 mm die with the frame and the device ID `QM2L-12Q`. 2. **Bond pads.** Two launch pads and twelve flux pads, on the 100 µm pad rule. 3. **Feedline.** A straight line between the launch pads. 4. **Qubits.** Twelve Xmons with trilayer SQUIDs. 5. **Readouts.** Twelve λ/4 hangers, one per qubit. 6. **Coupling.** None: a readout and flux test processor. 7. **Flux lines.** Twelve lines, each to its side pad. 8. **Whole chip.** Targets, one sim group per qubit, centring, save. ## LOM estimates Numbers are from the `.v1` notebook's saved output (the 12-qubit chip was not re-run for the docs): each qubit is solved with its readout and the shared feedline, at zero flux. The source states no qubit frequency; the drawn 15.31 nH SQUID puts every qubit at the same 3.258 GHz, below the PDK band of 4 to 8 GHz. | qubit | f_q (GHz) | α (MHz) | resonator | f_r (GHz) | κ (MHz) | χ (MHz) | |---|---|---|---|---|---|---| | Q01_L | 3.258 | -150 | R_Q01_L | 5.742 | 0.0657 | -0.12 | | Q01_R | 3.258 | -150 | R_Q01_R | 5.802 | 0.0677 | -0.117 | | Q02_L | 3.258 | -150 | R_Q02_L | 5.863 | 0.0698 | -0.115 | | Q02_R | 3.258 | -150 | R_Q02_R | 5.923 | 0.0718 | -0.113 | | Q03_L | 3.258 | -150 | R_Q03_L | 5.982 | 0.074 | -0.112 | | Q03_R | 3.258 | -150 | R_Q03_R | 6.042 | 0.0761 | -0.11 | | Q04_L | 3.258 | -150 | R_Q04_L | 6.102 | 0.0783 | -0.108 | | Q04_R | 3.258 | -150 | R_Q04_R | 6.161 | 0.0805 | -0.107 | | Q05_L | 3.258 | -150 | R_Q05_L | 6.221 | 0.0828 | -0.105 | | Q05_R | 3.258 | -150 | R_Q05_R | 6.28 | 0.0851 | -0.104 | | Q06_L | 3.258 | -150 | R_Q06_L | 6.34 | 0.0874 | -0.103 | | Q06_R | 3.258 | -150 | R_Q06_R | 6.399 | 0.0898 | -0.101 | These are lumped-model estimates (component formulas and calibration), not FEM results. [The LOM page](../../library/lom.md) says how far to trust them. ## Build spec The rest of this page is the detailed spec the notebook was built from, with every component, port, route and length. Source: [OJB-Quantum/Qiskit-Metal-to-Litho](https://github.com/OJB-Quantum/Qiskit-Metal-to-Litho), `12_Qubit_Flux_Tunable_Processor_in_Colab.ipynb` (O. J. Benally, 2026). Output GDS `GDS Files/twelve_qubit_flux_processor.gds` (not downloaded). Licence **CC BY 4.0** (`LICENSE` = Attribution 4.0 International; the notebook says "Open Access (CC-BY-4.0)"), verified. Read on 2026-11-02: notebook constants only, not run. What it is: 12 flux-tunable Xmons (`TransmonCrossFL`) in two columns of 6, one each side of a straight vertical feedline. Each has a λ/4 meander readout, frequency-multiplexed 6.05–6.82 GHz in 70 MHz steps, and a flux line fanned out to side pads. Quantum Metal draws the qubits and pads; PHIDL smooths the feedline and readout centrelines. Frame: chip 11 × 11 mm centred. **Box (0, 0, 11000, 11000)**, **Designer = source + (5500, 5500)**. Feedline CPW 11.7/5.1 µm; readout 10/6 µm (fillet 85); flux 6/8 µm tapering to 5/3 µm. Layers: 3 = metal, 5 = flux trace, 9 = junction. The chip needs `Xmon` with `flux_port`, `BondPad` and hangers. The repository also holds other designs (see bottom). ### Components | instance | catalog ref | key params (units) | at (x, y) Designer | rot | notes | |---|---|---|---|---|---| | Qnn_L (n = 1…6) | `Xmon` (`flux_port=True`) | cross tip-to-tip **520** (Metal cross_length 260), arm_width 35, gap 20. One claw toward the feedline: claw_length 190, width 10, gap 6, ground spacing 5 | x **3150**, y 8650 / 7390 / 6130 / 4870 / 3610 / 2350 | claw faces +x | Metal orientation −90; source x −2350, rows y 3150 … −3150 (pitch 1260) | | Qnn_R | same | same | x **7850**, same rows | claw faces −x | orientation 90 | | P_TOP / P_BOT | `BondPad` (pad 260 × 320, taper 260, neck 38, clearance 60) | | (5500, 10200) / (5500, 800) | 180 / 0 | launch pads at y ±4700 | | F_L1…6 / F_R1…6 | `BondPad` (260 × 150, taper 220, neck 42) | | x 650 (left) / 10350 (right), one per qubit | 270 / 90 | flux pads with edge x ±4850 | Junction: **L_J 15.904 nH** for all qubits (the SQuADDS tutorial-5 value), so `junction={"Lj_nH": 15.9, "Cj_fF": 2}` (Cj assumed). The SQUID is flux-tunable. ### Routes | route | from → to | type, length | hanger / waypoints | notes | |---|---|---|---|---| | FL | P_TOP.a → P_BOT.a | cpw 11.7/5.1, fillet 70 | straight x 5500, y 9700 → 1300 | travel −y, so left = +x (east column) | | R_k (12) | Qk.claw → FL.hanger | meander **L_k = v/(4 f_k)**, v = 1.20e8 m/s: 6.05 → 4959, 6.12 → 4902, 6.19 → 4847, 6.26 → 4792, 6.33 → 4739, 6.40 → 4688, 6.47 → 4637, 6.54 → 4587, 6.61 → 4539, 6.68 → 4491, 6.75 → 4444, 6.82 → 4399 µm | coupling section **260 µm**, 18 µm from the feedline (3 "teeth"; exact geometry is in the notebook's manual route code). s at each qubit row (s = 9700 − y_row), side right (west) / left (east) | meander amplitude solved per resonator (120–360 µm), leads 250 µm | | FLX_k (12) | Qk.flux → F_k.a | cpw 6/8 → 5/3 (taper 95 µm), fillet 55 | escape 360 µm outward, then to the side pad at its row | | Frequency assignment: 12 targets in site order (Q01_L, Q01_R, Q02_L, …), not checked against the rendered GDS. ### Launchers, pads, test structures, markers, labels 2 feedline pads + 12 flux pads. No fill in this notebook (the 6-Xmon notebook adds antidots). ### LOM node structure 12 grounded Xmon islands (SQUID, flux-tunable through FLX_k). Each claw gives C_g to R_k (λ/4), which gives κ to FL (two 50 Ω ports). There are no qubit–qubit couplers (readout and flux test processor). ### Targets Readout f_k 6.05–6.82 GHz (70 MHz spacing). The first-pass λ/4 model uses v_ph = 1.2e8 m/s. Qubit frequency is not stated. ### Simplifications for a first rebuild 1. Use our standard hanger (ground_gap ≈ 18 − 6 − 5.1 ≈ 7 µm, assumed meaning of the 18 µm gap) and skip the 3-teeth coupler. Flux lines are straight escapes plus one bend to the side pads. 2. Other repo content (reference only): `6_Xmon_Layout…` (spec [qm2litho_6xmon](qm2litho_6xmon.md)), a 400-transmon 20 mm EBL array (`GDS Files/400Q-20MM_*.gds`, `Python Code_BEAMER/400_Qubit_*.py`; ground contacts deliberately omitted), `Full Chip Ex-001.GDS`, `4 Qubit Demo.gds`, `Single Meander CPW Resonator Chip.gds`, `KI-TWPA_Example.gds`. These are GDS/DXF only and were not downloaded. ### Build The notebook uses `metal:TransmonCrossFL` (flux T-line options from the source), BondPad launch/flux pads with the source sizes, hangers 260 µm with 7 µm ground strip, flux lines straight at 6/8 µm.