Turning spin into a charge the sensor can read

Two hole spins sat in gate-defined quantum dots in the germanium device. The sensor does not observe spin directly. Pauli spin blockade first maps different spin states onto different charge states. That conversion is the first gear in the chain: a state mapped incorrectly remains a wrong answer even with a perfect sensor.[1]

The experiment showed that gate-ramp speed and magnetic-field direction alter the mapping. By separating the transitions, the team selected a regime that sends one spin state along the sole unblocked charge path. That let it trace loss at particular stages of preparation and readout, instead of placing one final percentage inside an opaque box.[1]

Making a short-lived signal last

Spin states decay quickly at the readout point. Their lifetimes shortened as magnetic field increased, so the final measurement used a lower field. A double latch then moved the spin difference into a charge difference that persisted longer. That charge state gave the radiofrequency sensor enough time to integrate a measurement.[1]

Average preparation-and-measurement fidelity for the two qubits reached 97.0%. Rabi-oscillation visibility and a confusion matrix tested the result. The earlier level of about 94% for germanium hole-spin readout is a useful reference, while device conditions, timing and field settings still govern how those numbers should be compared.[1]

The component a larger array must test

The instructive result is a separate control knob for each error path. Field and gate ramps shape spin-to-charge conversion; latching sets signal lifetime; the sensor contributes the remaining noise. A larger array would need those settings to work across many dots at once. Successful readout of this device has not yet measured that shared operating condition.[1]

The design is valuable because its result can be broken into physical steps. If field and gate settings vary among dots in a wider germanium array, the same readout sequence could perform differently. Array-scale measurements can reveal whether that variation is manageable. This paper carefully establishes the laboratory result on two qubits.[1]