Quantum Computer Qubits Linked by New Design for Faster, More Reliable Processing

Researchers are continually seeking improved methods for coupling superconducting qubits to build more complex and powerful quantum processors. Gihwan Kim, Andreas Butler, and Oskar Painter, from the Kavli Nanoscience Institute and the Institute for Quantum Information and Matter at the California Institute of Technology, demonstrate a novel cross-Kerr coupler design that circumvents limitations of traditional capacitive coupling schemes. Their approach utilises small Josephson energy SQUID couplers to achieve tunable interactions without relying on mode hybridization, thereby avoiding leakage errors and enabling faster gate speeds. This hybridization-free architecture represents a significant step towards realising densely packed, miniaturised superconducting quantum circuits with enhanced coherence and connectivity, and offers a promising pathway for scalable quantum computation. SQUID-based junction coupling enhances superconducting qubit control and fidelity by minimizing decoherence Researchers have developed a novel coupling architecture for superconducting qubits that addresses limitations in current designs. Conventional superconducting quantum circuits rely on capacitive charge-based linear coupling, which presents challenges in balancing qubit properties like anharmonicity, coherence, and connectivity. This work introduces a junction-based coupling scheme utilising superconducting quantum interference device (SQUID) couplers with small Josephson energies to overcome these constraints. The new design directly implements intrinsic cross-Kerr interactions, controlled by external fluxes, and crucially avoids reliance on mode hybridization, a common source of errors in existing systems. This innovative approach enables the implementation of a fast, adiabatic, and high-fidelity controlled-Z gate without introducing additional modes that could lead to leakage errors. Simulations demonstrate the operation’s robustness even with asymmetries in the coupler junctions, a common fabrication challenge, particularly for high-frequency qubits. By maintaining interactions at a perturbative scale, the SQUID couplers minimise unwanted mixing between coupled elements while achieving a strong cross-Kerr interaction originating from diagonal coupling elements. The research details how this architecture circumvents the trade-offs between dielectric loss and qubit size inherent in traditional capacitive coupling. Furthermore, the study addresses potential crosstalk issues arising from junction asymmetries and parasitic hybridization with neighbouring qubits, demonstrating these effects are sufficiently small for practical circuit parameters. Analysis reveals that coherent errors remain below 5x 10⁻⁷ for gate times of 22 nanoseconds, even with junction asymmetries up to 20 percent. This level of performance is critical for maintaining quantum information integrity during gate operations. As a demonstration of scalability, the researchers present a tiling strategy for a miniaturized superconducting quantum processor based on merged-element qubits, all interconnected via these junction-based SQUID couplers. This design paves the way for denser, more robust quantum processors with improved performance and reduced susceptibility to errors, representing a significant step towards practical quantum computation. SQUID coupler design for fast adiabatic controlled-Z gate implementation requires careful parameter optimization Superconducting quantum circuits employ SQUID couplers, superconducting quantum interference devices, with small Josephson energies to mediate qubit interactions. This junction-based coupling architecture addresses limitations inherent in traditional capacitive charge-based linear coupling schemes, which struggle to simultaneously maintain qubit anharmonicity, coherence, and connectivity. The research focused on designing couplers that avoid dynamical variations in mode hybridization, a source of non-adiabatic transitions and gate speed limitations. By utilizing SQUID couplers, the study aimed to achieve a fast, adiabatic, and high-fidelity controlled-Z gate without introducing extra modes into the system. Circuit quantization was used to numerically calculate the ZZ interaction rate, ζ13, as a function of junction asymmetry, ∆EJ,C, and to validate perturbative calculations.

The team assessed crosstalk effects by simulating a chain of three transmons connected by these SQUID couplers, revealing that junction asymmetries induce interactions between non-adjacent qubits. Energy level diagrams were constructed to visualise transition matrix elements arising from longitudinal interactions, and the ZZ interaction rate was calculated to quantify the strength of this crosstalk. To further investigate crosstalk, a circuit comprising two directly coupled transmons and a spectator transmon was analysed. Spectator ZZ interaction rates, ζ1S and ζ2S, were computed as functions of the spectator qubit’s transition frequency and coupler capacitance, Cpara, with Cpara set to 30 aF for initial analysis. These calculations demonstrated that the coherent error of a 22ns-long controlled-Z gate remained below the 10−6 level over a 1-hour drift period, even with static flux offsets of δΦe,i and δ(∆Φe,o) applied to the flux waveform. The study found that a junction asymmetry of less than 20% is sufficient to suppress ζ13/2π below 60kHz when ΣEJ,C/2π = 0.8GHz. High fidelity controlled-Z gate implementation via junction-based SQUID coupling and crosstalk analysis demonstrates improved performance metrics Researchers demonstrate a junction-based coupling architecture utilising superconducting quantum interference devices (SQUIDs) to achieve a fast, adiabatic, and high-fidelity controlled-Z gate without introducing extra modes. This work addresses limitations inherent in traditional capacitive charge-based linear coupling schemes commonly found in superconducting quantum circuits. The proposed SQUID couplers, featuring relatively small Josephson energies, maintain interaction at a perturbative scale, limiting undesired higher-order mixing between coupled qubits. Analysis reveals that the controlled-Z gate maintains a small coherent error below the 10−6 level over a 1-hour drift period. Junction asymmetries within the SQUID couplers can induce crosstalk between next-nearest-neighbor qubits, resulting from odd-parity interactions. Longitudinal interaction, calculated using second-order perturbation theory, demonstrates a ZZ interaction rate ζ13, which is in good agreement with numerical circuit quantization results. Calculations show that a junction asymmetry less than 20% is sufficient to suppress ζ13/2π below 60kHz for ΣEJ,C/2π = 0.8GHz. Parasitic hybridization between a transmon and a spectator qubit leads to an “indirect” ZZ crosstalk, ζ1S, which is maximized near the resonance condition ωS ≈ω2. With parameters including CS = 69.2 fF and Cpara = 30 aF, the study highlights the importance of avoiding resonant frequency collision and suppressing stray capacitances to levels at or below tens of aF. The research culminates in a fully miniaturized “mergemon” architecture, where qubits and coupling elements are implemented using Josephson junctions, eliminating bulky capacitor elements. This tiling strategy allows for increased qubit density within the array, envisioning the use of a consistent oxide barrier thickness for both SQUID couplers and high-frequency mergemons to simplify circuit fabrication. The proposed architecture features a common ground connected to all low-voltage-side junction electrodes, with high-voltage-side electrodes connecting each mergemon qubit to its four nearest-neighbors via a SQUID coupler. SQUID couplers enable controlled interactions and minimise unwanted qubit mixing, leading to improved quantum circuit performance Superconducting quantum circuits commonly employ capacitive charge-based linear coupling to mediate interactions between qubits. This established method presents difficulties in simultaneously achieving large qubit anharmonicity, coherence, high connectivity, and compact circuit layouts. Tunable interactions facilitated by linear coupling can also generate dynamic variations in mode hybridization, potentially leading to non-adiabatic transitions and limiting the speed of quantum gate operations. Researchers have now proposed a junction-based coupling architecture utilising superconducting quantum interference devices (SQUIDs) with small Josephson energies to address these limitations. SQUID couplers intrinsically provide cross-Kerr interactions controllable via external fluxes, circumventing the need for mode hybridization. The small Josephson energies of these couplers maintain interactions at a perturbative scale, minimising unwanted higher-order mixing between qubits while still enabling a strong cross-Kerr interaction originating from diagonal coupling elements. This approach facilitates the implementation of a fast, adiabatic, and high-fidelity controlled-Z gate without introducing additional modes of operation, demonstrating robustness against junction asymmetry for high-frequency qubits. Although unconventional crosstalk may occur due to junction asymmetries and parasitic hybridization, analyses indicate these effects are sufficiently small for practical circuit parameters. A scalable tiling strategy for miniaturised superconducting quantum processors based on merged-element qubits has also been demonstrated using this junction-based coupling. The development of this SQUID coupler offers modeless and tunable cross-Kerr coupling, realising a fast and high-fidelity controlled-Z gate between transmons with minimal adiabaticity overhead. Sensitivities to junction asymmetry and flux noise are suppressed by utilising high qubit frequencies and relatively small junction energies. The findings demonstrate that unconventional crosstalk, arising from parasitic hybridization and junction asymmetry, remains sufficiently small for realistic circuit parameters. This junction-based coupling scheme is particularly suited to merged-mon qubit architectures, anticipating junctions with small Josephson energies and reduced SQUID loop sizes due to the absence of bulky shunt capacitors, and ultimately enabling a scalable tiling strategy towards fully miniaturised superconducting quantum processors. 👉 More information 🗞 A Tunable, Modeless, and Hybridization-free Cross-Kerr Coupler for Miniaturized Superconducting Qubits 🧠 ArXiv: https://arxiv.org/abs/2602.03186 Tags: