Superconducting dual rail cavity qubits and error correction*

            I will present a new architecture for bosonic error correction (https://doi.org/10.1073/pnas.2221736120) which adapts and extends the well known “dual-rail” encoding, where a single photon can be superposed between two distinguishable microwave cavity modes. When the techniques of circuit QED are applied to this system, we can develop a full family of universal gates ( https://doi.org/10.1103/PRXQuantum.4.020354) in which all of the known decoherence errors can be efficiently detected, allowing postselected gate fidelities higher than any other solid-state platform. Recent experimental breakthroughs (https://arxiv.org/abs/2212.11929) on this type of quantum hardware have dramatically improved our ability to perform high-fidelity single qubit operations in this system, with error rates at a few parts in 10^5. Because all operations, including single and two-qubit gates as well as state preparation and measurement (SPAM), allow any single error to be detected and classified as an erasure, we can achieve performance that is orders of magnitude better than other superconducting architectures. I will present results (https://arxiv.org/abs/2307.03169) showing SPAM that can attain overall error rates of a few parts in 10^4, exceeding all other quantum computing platforms. In addition, these experiments confirm the highly favorable hierarchy of errors in the cavity dual-rail, with intrinsic dephasing rates of several milliseconds. This new encoding and error hierarchy, in which the dominant errors can be converted to erasures, is expected to achieve performance well past the thresholds for higher level codes such as the surface code, allowing a faster path to fault-tolerant and scalable computing. I will review the latest results with this new platform, including the progress on multi-qubit operations.