Measuring error rates of mid-circuit measurements

Area of Science:

• Quantum Information Science
• Quantum Computing Hardware
• Quantum Error Correction

Background:

• High-fidelity mid-circuit measurements are essential for advancing quantum computing, enabling fault tolerance and complex computations.
• Existing methods for evaluating mid-circuit measurement performance are limited, creating a gap in comprehensive assessment.
• Measurement-induced crosstalk can introduce errors, degrading the performance of multi-qubit processors.

Purpose of the Study:

• To introduce the first randomized benchmarking protocol specifically designed to measure errors induced by mid-circuit measurements in many-qubit circuits.
• To address the lack of comprehensive performance assessment methods for mid-circuit measurements.
• To detect and mitigate previously uncharacterized measurement-induced crosstalk.

Main Methods:

• Development and application of a novel randomized benchmarking protocol to quantify measurement-induced error rates.
• Utilizing the protocol to identify and eliminate crosstalk errors on a 20-qubit trapped-ion quantum computer.
• Applying the protocol to a 27-qubit IBM Q processor to measure crosstalk error and its mitigation by dynamical decoupling.

Main Results:

• Successfully detected and eliminated previously undetected measurement-induced crosstalk in a 20-qubit trapped-ion quantum computer.
• Quantified the rate of measurement-induced crosstalk error on a 27-qubit IBM Q processor.
• Demonstrated the effectiveness of dynamical decoupling in reducing measurement-induced crosstalk error.

Conclusions:

• The developed randomized benchmarking protocol provides a comprehensive method for assessing mid-circuit measurement performance.
• This protocol is crucial for improving the fidelity of quantum computations and enabling scalable quantum computing.
• The findings highlight the importance of addressing measurement-induced crosstalk for robust quantum error correction and computation.