OXFORD, UK, March 21, 2025 -- The University of Oxford research group led by OQC CSO Dr. Peter Leek today announced research demonstrating a fundamental speedup of the controlled-Z gate in superconducting qubits reaching a fidelity of 99.8% in only 25 ns.

This new approach, presented by Dr. Simone Fasciati at the American Physical Society’s Global Physics Summit (APS) in Anaheim, employs a new twist on the coaxmon circuit architecture that simultaneously eliminates a leading source of error and accesses a triple-state degeneracy to further speed up one of the fastest existing two-qubit (2Q) gates in quantum computing, widely used in superconducting circuits.

The fidelity of 2Q gates is a key metric for building hardware towards useful quantum computers. The further the fidelity is above a fault-tolerance threshold (typically 99–99.5%), the more efficiently remaining errors can be corrected using quantum error correction. Another key requirement is that gates are fast — as well as enabling high fidelities via reduced decoherence, the gate speed determines how fast a useful quantum computation can be performed. This breakthrough work achieves this high fidelity and speed with a particularly simple circuit design, requiring less hardware complexity than leading 2Q gate implementations based on tunable couplers.

“It’s one of the fastest 2Q gates ever benchmarked, has a very competitive gate fidelity, and one of the simplest circuit designs.” says OQC CSO, Dr. Peter Leek, “That combination bodes extremely well for scaling up to perform fast and useful quantum computing.”

Improved gates for superconducting qubits

Commercially useful quantum computation requires fast and high-fidelity entangling gates between qubits. In superconducting quantum circuits, many common realisations of the two-qubit controlled-Z (CZ) gate involve an interaction between a pair of states in the two-excitation subspace.

Research has proposed theoretically that increasing the number of states involved in performing the gate from two to three would increase its speed by  a factor of √2. This scheme requires two qubits with exactly opposite ‘anharmonicities’, a property of the energy levels of superconducting qubits. The team at Dr. Leek’s research lab has experimentally demonstrated such a system by coupling a standard transmon with a specially-designed qubit of opposite anharmonicity, an inductively shunted transmon. In their previous research, they showed that this combination of qubit species suppresses ZZ idling error, a leading source of error  in superconducting qubits. Typically it is suppressed by adding tunable couplers between qubits, which adds  complexity to the hardware. Here it is instead sufficient to use a fixed coupling architecture, keeping complexity to a minimum.

The team was able to observe the expected √2 speed-up and to execute a CZ gate in less than 25ns and with a fidelity of 99.8%. Presented at the APS Global Physics Summit by Dr. Simone D Fasciati from the University of Oxford within the ‘Superconducting Qubits: Improved Gates’ session, the experiments have contributed to exciting progress for superconducting qubits.

“The idea behind this gate speed-up is simple and elegant: to make more efficient use of the computational resources already available in a two-qubit system. Perhaps the same principle could be applied to unlock more performance in other existing circuits”, explained Fasciati.

The significance for commercial superconducting quantum computers

Achieving this world-leading combination of high-fidelity and fast 2Q gates validates the ongoing progress of superconducting qubits towards improving the accuracy, reliability and speed of quantum computations.

The next challenge is then to apply these academic research advances at scale. Traditionally, 2D circuits require increasingly intricate engineering to route control wiring across the chip to the qubit which degrades the quality of the qubits, and can increase engineering errors. Similarly, a key challenge of superconducting circuits for quantum computation is the ability to scale qubit numbers whilst maintaining qubit quality and control in order to reach a commercially useful level of processing power.

The latest work builds on the simplicity of the ‘coaxmon’ architecture — one of the world’s simplest three-dimensional quantum architectures that bring key componentry off-chip which enables improved flexibility, engineerability, and scalability. This patented technology is already being used in commercial environments by OQC. Further developing the coaxmon, the new work demonstrates the key properties of superconducting qubits required for reliable quantum computing, without adding the additional hardware complexity that is usually required for suppressing the  ZZ error. A patent has been filed by the Oxford research team on the technique for realising the opposite anharmonicity required for this fast gate and error suppression.

OQC has a history of building on academic breakthroughs with further research and engineering, to advance next-generation quantum computers. For instance, OQC advanced previous Leek Lab research on cross-resonance 2Q gates, engineering the gates to achieve significantly higher fidelities on larger scales, now used across OQC devices. The 2Q gate results therefore serve as a proof-of-concept demonstrator for further development and optimisation in OQC’s coaxmon-based devices. This means that the new research can be directly put towards achieving the hardware metrics needed for commercially useful quantum computers.