Dynamically corrected gates in silicon singlet-triplet spin qubits

TL;DR

This paper demonstrates the first experimental implementation of dynamically corrected gates in silicon singlet-triplet qubits, significantly reducing gate errors and achieving fidelities above 0.99, thus advancing fault-tolerant quantum computing.

Contribution

It introduces experimentally realized dynamically corrected gates in silicon singlet-triplet qubits, addressing control constraints and noise mitigation in semiconductor quantum dots.

Findings

Gate infidelities reduced by a factor of three

Achieved gate fidelities above 0.99

Identified unexpected pulse distortions affecting performance

Abstract

Fault-tolerant quantum computation requires low physical-qubit gate errors. Many approaches exist to reduce gate errors, including both hardware- and control-optimization strategies. Dynamically corrected gates are designed to cancel specific errors and offer the potential for high-fidelity gates, but they have yet to be implemented in singlet-triplet spin qubits in semiconductor quantum dots, due in part to the stringent control constraints in these systems. In this work, we experimentally implement dynamically corrected gates designed to mitigate hyperfine noise in a singlet-triplet qubit realized in a Si/SiGe double quantum dot. The corrected gates reduce infidelities by about a factor of three, resulting in gate fidelities above 0.99 for both identity and Hadamard gates. The gate performances depend sensitively on pulse distortions, and their specific performance reveals an unexpected distortion in our experimental setup.