XMonsDirectCoupling (KQC)

A KQCircuits chip with three Swissmon qubits in a row. Neighbouring qubits are coupled directly: their arm tips face each other across a small gap. Each qubit has a T-tapped quarter-wave readout into one probe line. It is rebuilt on the Quantcore PDK.

Build it to see direct qubit-to-qubit coupling (declared in the netlist with d.couple) and T-tapped readouts.

The whole xmonsdirectcoupling chip as the notebook builds it

At a glance

Notebook

examples/designs/kqc/xmonsdirectcoupling.ipynb (an older one-cell version is kept as xmonsdirectcoupling.v1.ipynb)

Source

KQCircuits chips/xmons_direct_coupling.py (XMonsDirectCoupling), © IQM Finland Oy, GPL-3.0

Chip

10 × 10 mm, SMA8 launchers, chip name “XMDC”

On the chip

3 Swissmons 350 µm apart · 3 T-tapped λ/4 readouts · 1 probe line · 3 charge and 3 flux lines · 4 junction-test arrays

Builds today

yes (layout built on 2026-11-19 in 8 s)

How the notebook builds it

The layout segments of XMonsDirectCoupling, one picture each
  1. Die. The empty 10 × 10 mm chip with the Quantcore frame (dicing marks, alignment grids, chip boundary, device ID) and the eight launchers.

  2. Bond pads. The launchers moved so their outer etch edge sits 100 µm from the die edge (the pad rule).

  3. Qubits. Three Swissmons in a row with Quantcore SQUIDs.

  4. Readout line. A T, capacitor and tail per qubit, then the probe line.

  5. Readouts. The T-tapped λ/4 resonators.

  6. Direct coupling. Declared with d.couple (no part of the layout joins the qubits).

  7. Control lines. Charge and flux lines.

  8. Whole chip. Sim groups, centring, design saved.

LOM estimates

Numbers are from a run of this notebook on 2026-11-19 (Setup, Design, Check and the LOM cells, 10 s): QuCat normal modes with QuTiP dressed values.

group

pair

ZZ (kHz)

neighbours

hybridised

QB1–QB2–QB3

QB1-QB3

–

no

yes

QB1–QB2–QB3

QB1-QB2

–

yes

yes

QB1–QB2–QB3

QB2-QB3

–

yes

yes

group

qubit

f_q (GHz)

α (MHz)

resonator

f_r (GHz)

κ (MHz)

χ (MHz)

QB1+RR1

QB1

6.521

-255

RR1

5.204

10.2

-2.3

QB2+RR2

QB2

6.543

-256

RR2

5.418

11.6

-2.93

QB3+RR3

QB3

6.522

-254

RR3

5.376

10.4

-3.46

LOM plot of xmonsdirectcoupling

These are lumped-model estimates (component formulas and calibration), not FEM results; see the LOM page for how far to trust them.

Build spec

The rest of this page is the detailed spec the notebook was built from: every component, port, route and length.

Source other-libs/KQCircuits-main/KQCircuits-main/klayout_package/python/kqcircuits/chips/xmons_direct_coupling.py (XMonsDirectCoupling), GPL-3.0 (IQM Finland). Rebuilt from extracted parameters. What it is: 3 flux-tunable Swissmons (“Finnmons”) in a row, 350 µm apart, so neighbouring arm tips face each other across 10 µm and couple directly. Each qubit has a T-tapped λ/4 readout resonator (86 Ω, a=5 b=20) coupled through a finger capacitor to one probeline, which has a series input capacitor. There is a drive line and a flux line per qubit, plus the standard junction-test arrays. 10 × 10 mm, KQC frame = Designer frame, face 1t1. Checked on 2026-11-02: every length matches KQC, and the LOM has no unconnected ports.

Components

instance

catalog ref

key params (units)

at (x, y)

rot

notes

QB1

Swissmon

COM + cpl_gap=[110,90,110], cpl_length=[0,158,0], cpl_width=[60,24,60], cl_offset=[150,-150]

(4650, 5000)

0

cplr1 N (4650,5209); drive (4500,5150); flux S (4650,4746.67)

QB2

Swissmon

COM + cpl_gap=[110,102,110], cpl_length=[0,140,0], cpl_width=[60,24,60], cl_offset=[150,150]

(5000, 5000)

0

cplr1 (5000,5215); drive (4850,4850); flux (5000,4746.67)

QB3

Swissmon

as QB1 but cl_offset=[-150,-150]

(5350, 5000)

0

cplr1 (5350,5209); drive (5500,5150); flux (5350,4746.67)

PL1–PL3

WaveguideCoplanarSplitter

lengths=[11,11,31], port_names=["right","left","bottom"]

(x_k, 7900), x = 4650/5000/5350

0

probeline T’s; bottom faces down

CK1–CK3

FingerCapacitorSquare

finger_number=4, finger_length 60.8 / 61.55 / 58.1, CAP

snap b → PLk.bottom

(→ 90)

κ capacitor

TP1–TP3

WaveguideCoplanarTaper

a=5, b=20, a2=10, b2=6, taper_length=100

snap b → CKk.a

(→ 90)

50 → 86 Ω

RT1–RT3

WaveguideCoplanarSplitter

lengths=[22.5,22.5,62.5], a_list=[5,5,5], b_list=[20,20,20], port_names as PL

snap right → TPk.a

(→ 90)

tap T in the resonator: left goes down to the qubit, bottom goes east to the tail

IN

FingerCapacitorSquare

finger_number=8, finger_length=55.4, CAP

(4294.8, 7900)

0

series input cap (“35 fF” in source); a (4239.6,7900), b (4350,7900)

testarray_w/s/e/n

JunctionTestPadsSimple

as in single_xmons

(350,3750) / (3750,350) / (8350,3750) / (3750,8350)

0

standard KQC junction tests

COM = dict(arm_length=[146]*4, island_r=2, port_width=[10,5,10], arm_width=[24]*4, gap_width=[24]*4); CAP = dict(finger_gap=5, finger_gap_end=5, finger_width=15, ground_padding=10).

Routes

route

from → to

type, length (µm)

waypoints

notes

RR1 / RR2 / RR3

RTk.left → QBk.cplr1

cpw a=5 b=20; 2399.2 / 2392.45 / 2401.9

none (straight down)

KQC: n_bridges=10

RS1 / RS2 / RS3

RTk.bottom → free_end((x_k+162.5, y_k), term=0)

cpw a=5 b=20; 2384.4 / 2204.85 / 2199.8

none (stub east 100, then down to y = 5303.38 / 5482.18 / 5490.68)

shorted tail of the λ/4

PL_in

chip.NW → IN.a

cpw 2696.68

(2800, 7900)

PL_a / PL_b / PL_c

IN.b → PL1.left / PL1.right → PL2.left / PL2.right → PL3.left

289 / 328 / 328

none

PL_out

PL3.right → chip.NE

3096.08

(7200, 7900)

DL1

chip.WN → QB1.drive

cpw term2=6, 4559.48

(2500, 7200)

diagonal

DL2

chip.SW → QB2.drive

term2=6, 4890.53

(2800,2500), (4850,4550)

DL3

chip.EN → QB3.drive

term2=6, 4559.48

(7500, 7200)

FL1

chip.WS → QB1.flux

cpw 4366.09

(900,2800), (4650,4646.667)

diagonal

FL2

chip.SE → QB2.flux

4864.76

(7200,2500), (5000,4646.667)

FL3

chip.ES → QB3.flux

4366.09

(9100,2800), (5350,4646.667)

d = Designer("kqc_xmons_dc", launchers="SMA8", chip={"name_chip": "XMonsDC"})
q1 = d.add("QB1", "Swissmon", at=(4650, 5000), junction=J, cpl_gap=[110, 90, 110], cpl_length=[0, 158, 0],
           cpl_width=[60, 24, 60], cl_offset=[150, -150], **COM)          # QB2, QB3 likewise
for k, x, lf in ((1, 4650, 60.8), (2, 5000, 61.55), (3, 5350, 58.1)):
    d.add(f"PL{k}", "WaveguideCoplanarSplitter", at=(x, 7900), lengths=[11, 11, 31], port_names=["right", "left", "bottom"])
    d.add(f"CK{k}", "FingerCapacitorSquare", snap="b", to=f"PL{k}.bottom", finger_number=4, finger_length=lf, **CAP)
    d.add(f"TP{k}", "WaveguideCoplanarTaper", snap="b", to=f"CK{k}.a", a=5, b=20, a2=10, b2=6, taper_length=100)
    d.add(f"RT{k}", "WaveguideCoplanarSplitter", snap="right", to=f"TP{k}.a", lengths=[22.5, 22.5, 62.5],
          a_list=[5, 5, 5], b_list=[20, 20, 20], port_names=["right", "left", "bottom"])
    d.route(f"RR{k}", f"RT{k}.left", f"QB{k}.cplr1", a=5, b=20)
    d.route(f"RS{k}", f"RT{k}.bottom", Designer.free_end(at=(x + 162.5, {1: 5303.38, 2: 5482.18, 3: 5490.68}[k]), term=0), a=5, b=20)
d.couple("C12", "QB1", "QB2"); d.couple("C23", "QB2", "QB3")     # direct arm-tip coupling (value from a C-matrix)

Launchers, pads, test structures, markers, labels

SMA8 default names, all 8 used: NW = probeline in, NE = probeline out; WN/SW/EN = drive QB1/QB2/QB3; WS/SE/ES = flux QB1/QB2/QB3. There are 4 junction-test arrays.

LOM node structure (qubit ↔ resonator ↔ feedline ↔ bus ↔ control lines)

  • QB1–QB3: island + SQUID + cplr1. QB1–QB2 and QB2–QB3 are direct capacitances via d.couple (C-matrix value, sim_group over QB1..QB3).

  • Readout k: QBk.cplr1 ≡ RTk node (the RR route is a wire). RSk becomes a λ/4 TLres there. CKk is the series C (estimate ≈ 20 fF) to the probeline node (PL1–PL3 merged). IN is the series C (estimate 39.8 fF) from chip.NW to the probeline, and chip.NE is a 50 Ω port.

  • LOM gap: the real resonator is a tapped λ/4, RR (2.4 mm) + T + RS (2.2–2.4 mm) ≈ 4.7–4.9 mm. The builder only counts RS as the TLres (lumped f ≈ 13–14 GHz) and treats RR as a wire. For a correct LOM, override the length (KQC length = 4868.6 / 4682.3 / 4686.7) or add a “resonator through splitter” rule.

  • DLk becomes a drive control and FLk a flux_bias control on J_QBk.

Targets (if stated)

“Target parameters V2”: resonator lengths 2421.5+2447.1, 2436+2246.3, 2438.5+2248.2 µm (non-tail + tail), finger lengths 60.8 / 61.55 / 58.1 µm, input C ≈ 35 fF. Chip params: qubit_spacing=10, arm_width_a=arm_width_b=24, rr_cpl_width=[24,24,24].

Simplifications for a first rebuild

Skip the junction tests and airbridges. KQC puts n_bridges of 25/30 on the control lines, 10/5 on the resonators and 1–4 on the probeline. If a single conductor per resonator is wanted in the LOM, keep the RT/TP chain in the layout but set the TLres length by hand until the builder supports tapped resonators.

Quantcore PDK rebuild

The notebook is a Quantcore-PDK rebuild of the source chip. The netlist, LOM and imported components are kept, and the process is Quantcore’s.

Notebook: examples/designs/kqc/xmonsdirectcoupling.ipynb.

  • Chip: the Quantcore die frame (L dicing markers 1000/0, 3 × 3 alignment grids 1001/0, chip boundary 100/0, device ID where an edge is free; any brand text is QCT), drawn on the source’s 10 × 10 mm chip (its launchers sit on the chip edges, so the design does not fit the 7.5 mm die). KQC’s ChipFrame (dicing edge, corner markers, labels) is not drawn. KQC’s original grids its whole ground plane (with_grid, 5 µm squares on a 10 µm step). The rebuild keeps that coverage at the PDK density instead (chip parameter cheese: "chip", with_grid: false): 5 µm squares with 20 µm gaps over the whole die, at least 4 µm from the 3/0 keep-outs, inside the design area and clear of the frame markers and the device ID.

  • Stack: sapphire 650 µm, εr 10.25 (the PDK default stack, used by the LOM estimates and every FEM export); KQC’s own simulation default is silicon (εr 11.45), so LOM capacitances and resonator frequencies here are for sapphire.

  • Junctions: Quantcore Nb/AlOx/Nb trilayer junctions (junction_type from layouts.registry.trilayer_junction): fingers on 2/0 + 4/0, spacer 50/0 with the 51–53/0 gradient, sized on the PDK junction rule (Nb Δ = 1.4 meV, I_c·R_n = πΔ/2e, R_n·A 13.5 kΩ·µm² and C_J 50 fF/µm² assumed); every Swissmon’s SQUID becomes a PDK SQUID of two 0.25 × 0.6 µm junctions → R_n = 45 kΩ in parallel, L_J = 6.73 nH at zero flux, C_J = 15 fF (in the 40–50 kΩ window). KQC’s junction_width sets the straight finger. This replaces the assumed Lj 16 nH + 2 fF. Expected f_q ≈ 6.5 GHz (was 4.47–4.49 GHz), above the ≈ 5.3–5.5 GHz readouts; the three qubits stay identical and hybridised (estimate from the pre-rebuild LOM C_Σ). The four JunctionTestPads arrays get the same Quantcore SQUID.

  • Kept: the topology and every KQC element with the source’s parameters, positions and orientations, the launcher assignment, every route’s path and length, the sim groups (the T-tapped readouts with KQC’s tail lengths, the assumed 0.3 fF direct coupling).