论文标题
基准测试平面蜂窝代码
Benchmarking the Planar Honeycomb Code
论文作者
论文摘要
我们通过描述不需要额外的物理连接的边界以及优化量子贴片的形状来改善平面蜂窝代码。然后，我们使用Monte Carlo采样对代码进行基准测试以估计逻辑错误率，并得出包括阈值，Lambdas和Teraquop Qubit计数在内的指标。我们确定平面蜂窝代码可以使用7000个物理Qubits以0.1％的门级错误率（或给定的900个物理量子台（给定给定给定的本机两分之二奇偶校验测量）），使用7000个物理Qubits创建逻辑量子量子。我们的结果巩固了Honeycomb代码，作为二维量子架构的有前途的候选人，连通性很少。
We improve the planar honeycomb code by describing boundaries that need no additional physical connectivity, and by optimizing the shape of the qubit patch. We then benchmark the code using Monte Carlo sampling to estimate logical error rates and derive metrics including thresholds, lambdas, and teraquop qubit counts. We determine that the planar honeycomb code can create a logical qubit with one-in-a-trillion logical error rates using 7000 physical qubits at a 0.1% gate-level error rate (or 900 physical qubits given native two-qubit parity measurements). Our results cement the honeycomb code as a promising candidate for two-dimensional qubit architectures with sparse connectivity.
