Curator's Take
This article demonstrates a 180 ns CZ gate on protected multimode “P‑mon” qubits that reaches 99.62 % fidelity while suppressing idle‑state ZZ crosstalk to under 4 kHz, a combination rarely achieved in superconducting hardware. By using separate mediator modes for coupling, the work sidesteps the decoherence trade‑offs that have limited previous two‑qubit gates and moves the platform closer to the error‑correction thresholds needed for large‑scale processors. The result shows a clear path toward scaling superconducting arrays without sacrificing coherence, though integration with full control stacks and multi‑gate sequences will be the next hurdle.
— Mark Eatherly
Summary
To scale superconducting quantum processors, it is essential to achieve long coherence times while engineering interactions that do not introduce additional decoherence channels. In superconducting qubit systems, this can be realized using multimode circuits that feature a protected qubit mode alongside a distinct mediator mode. Building on this concept, our recently developed P-mon qubit provides intrinsic protection against decoherence from the readout environment. We extend this approach to controlled two-qubit interactions, by exploiting the mediator modes of P-mons for on-demand coupling. Because direct interactions between the qubit modes are strongly suppressed, unwanted $ZZ$-type interactions are significantly reduced to below $3.6(5)~\text{kHz}$ in the idle state. When tuning the coupled mediator modes on resonance, the cross-Kerr interaction between the qubit and the hybridized mediator modes leads to a qubit-state dependent frequency shift. By selectively addressing these transitions, we implement a $180~\text{ns}$ long CZ gate and determine a fidelity of $99.62(4)~\text{%}$. These results represent a significant step toward a scalable superconducting architecture that maintains high performance at scale.