Noise-resistant quantum state compression readout

Qubit measurement is generally the most error-prone operation that degrades performance of near-term quantum devices, and exponential decay readout fidelity severely impedes development large-scale information processing. Given these disadvantages, we present a state method, named compression readout, naturally avoids large multi-qubit errors by compressing into single qubit for measurement. Our method outperforms direct measurements in terms accuracy, advantage grows with system size. Moreover, because only one-qubit are performed, our requires solely fine calibration on one free correlated error, which drastically diminishes demand device calibration. These advantages suggest can immediately boost devices will greatly benefit computing.