State preparation fidelity by QND readout of a silicon electron spin qubit*

A quantum non-demolition (QND) single-shot qubit readout is crucial for large-scale quantum computing, and unlike conventional counterparts, can serve as a state preparation process with its error extinguishable by repetition. Here we discuss the state preparation fidelity of a cumulative QND readout of an electron spin qubit in a silicon quantum dot with its neighboring electron spin used as an ancilla [1]. By repeating the single QND readout process, a qubit spin state is prepared in the state correlated to the readout outcomes with a state dependent fidelity. Furthermore, a series of readout outcomes can be converted to the likelihood ratio of successful state preparation, which allows one to select events with a higher preparation fidelity. Combined with real-time signal processing, such protocols offer alternative routes for high-fidelity state preparation.

[1] J. Yoneda et al, Nat. Commun. 11, 1144 (2020).