Fast and Accurate Decoder for the XZZX code Using Simulated Annealing

Abstract: The XZZX code is a variant of the surface code designed to address biased noise in realistic quantum devices. For the XZZX code, we propose a decoder based on simulated annealing (SA). Our SA decoder can be readily and efficiently parallelized, by virtue of its simple MCMC-based algorithm. To prepare an initial configuration of SA, we propose to employ recovery chains obtained by a decoder which utilizes a kind of greedy matching graph algorithm. Although $Z$-biased noise is commonly assumed in real quantum devices, we focus on $Y$-biased noise, for which the minimum-weight perfect matching (MWPM) algorithm fails to decode accurately. Our numerical simulation for the code capacity noise model, where only data qubits suffer errors, confirmed that our SA decoder is more accurate than the MWPM decoder. Furthermore, our SA decoder attained the accuracy equivalent to that of the optimal decoder formulated by integer programming, called CPLEX decoder. In our greedy matching decoder, we randomly determine order of matching pairs of incorrect syndromes that have the same distance. This randomness brings about a variety of initial configurations of SA, which leads to faster convergence of our SA decoder. By comparing decoding times of our SA decoder, the CPLEX decoder, and matrix product state (MPS) decoder, all of which can handle $Y$-biased noise appropriately, we confirmed that our SA decoder is fastest if parallelized. This result implies a potential for combining of our greedy matching and SA decoder for practical use in quantum computing.