Gmon

Figure

Description

The Gmon is a superconducting qubit architecture developed at Google that adds a tunable coupler between neighboring Xmon-style qubits. Introduced by Y. Chen et al. (2014), the “g” in Gmon refers to the tunable coupling strength $g$ between qubits.

The tunable coupler is itself a frequency-tunable transmon-like element placed between two computational qubits. By adjusting its frequency via a flux line, the effective qubit-qubit coupling can be tuned from a finite positive value through zero to a finite negative value. This enables:

1. Fast two-qubit gates: bringing qubits into resonance with strong coupling for $i$SWAP-like gates in ~10–20 ns.
2. Idle isolation: parking the coupler to cancel residual $ZZ$ coupling, suppressing always-on errors.
3. CZ gates: diabatic flux pulses that accumulate a conditional phase.

The Gmon/tunable-coupler architecture was used in Google’s Sycamore (2019) and Willow (2024) processors and has become the dominant paradigm for frequency-tunable superconducting qubit arrays.

Hamiltonian

The three-body system (qubit 1 — coupler — qubit 2):

$H={\sum }_{i=1,2,c}{\omega }_i{a}_i^{†}{a}_i+\frac{{\alpha }_i}{2}{a}_i^{†}{a}_i^{†}{a}_i{a}_i+{\sum }_{i<j}{g}_{ij}(a_i^{†}{a}_j+a_i{a}_j^{†})$

The effective qubit-qubit coupling after adiabatically eliminating the coupler:

$g_{\text{eff}}=g_{12}-\frac{g_{1c}{g}_{2c}}{{\Delta }_c}$

where ${\Delta }_c={\omega }_c-({\omega }_1+{\omega }_2)/2$ is the coupler detuning. Setting $g_{12}=g_{1c}{g}_{2c}/{\Delta }_c$ achieves zero effective coupling (idle point).

Motivation

Fixed-coupling architectures suffer from always-on $ZZ$ interaction, causing idle errors and frequency-crowding constraints. The tunable coupler solves both: it allows fast gates when coupling is “on” and near-perfect isolation when “off,” dramatically improving circuit fidelity for multi-qubit algorithms.

Experimental Status

First demonstration — Y. Chen et al. (2014):

• Demonstrated tunable coupling between two Xmon-style qubits via a flux-tunable coupler element.
• Achieved two-qubit gate fidelities suitable for quantum error correction exploration.
• Showed ON/OFF switching of effective coupling by tuning the coupler frequency.

Sycamore processor — Arute et al. (2019):

• 53-qubit processor using Gmon-style tunable couplers for 86 qubit pairs.
• Median CZ fidelity 99.4% via randomized benchmarking.
• Demonstrated quantum computational advantage.

Willow processor — Google Quantum AI (2025):

• Improved to 99.7–99.85% median CZ fidelity.
• Enabled below-threshold surface code operation with 105 qubits.

Key Metrics

Metric	Value	Notes	Fidelity reference
2Q gate fidelity	99.5–99.9%	CZ or $i$SWAP	Y. Chen et al. 2014
2Q gate time	10–30 ns	Fast parametric gates	—
Residual $ZZ$ (off)	<10 kHz	At idle point	—
1Q gate fidelity	99.9%+	Same as Xmon	Y. Chen et al. 2014
Operating temperature	10–20 mK	Dilution refrigerator	—

References

Original proposal and demonstration

• Y. Chen et al., “Qubit Architecture with High Coherence and Fast Tunable Coupling,” Phys. Rev. Lett. 113, 220502 (2014)

Key experimental milestones

• F. Arute et al., “Quantum supremacy using a programmable superconducting processor,” Nature 574, 505 (2019)
• Google Quantum AI, “Quantum error correction below the surface code threshold,” Nature 638, 920 (2025)

Linked Papers

• chen-2014-gmon

Related Entries

• xmon — predecessor qubit architecture without tunable coupling
• transmon — parent qubit family
• tunable-coupler — the coupling element technology in detail
• circuit-qed — measurement and readout framework