Calibration method reduces errors in simultaneous silicon qubit control
A five-qubit experiment maintained approximately 99.9% accuracy when all qubits operated together, researchers reported in Nature Physics.
Researchers at the RIKEN Center for Emergent Matter Science and collaborating institutions have demonstrated a way to reduce errors when controlling several silicon spin qubits simultaneously. Their experiment achieved fidelity above 99.99% with up to three qubits active together, according to a study published in Nature Physics.
The team tested five qubits linked to one microwave control line. Individually, every qubit recorded single-qubit gate fidelity well above 99.99%, and the best result approached 99.999%. Fidelity measures how closely an operation follows its intended outcome. Operating all five together brought accuracy to approximately 99.9%.
The researchers identified the AC Stark shift as a major contributor to errors during parallel operation. Microwave signals can change a qubit’s effective energy levels, shifting the phase of its quantum state and interfering with the operation being performed.
Their calibration approach uses measurements between pairs of qubits to compensate for those shifts, avoiding a separate calibration for every possible combination. The experiments also incorporated changes to the device design and adjustments to microwave pulses.
Errors remained, however. The team linked the residual loss of accuracy to drive-induced decoherence: the higher microwave power needed for simultaneous control disrupted quantum coherence. Leon Camenzind, a RIKEN research scientist who supervised the experiments, said further device-design improvements could help reduce this limitation.
The researchers said the findings support shared control lines as an approach to expanding silicon quantum processors. Silicon spin qubits can be made using semiconductor manufacturing methods already used for computer chips.
