Extended Data Fig. 3: Two-qubit couplings.
All measurements are performed with all qubits initialized to spin-down. a, Schematic sequence of the exchange spectroscopy measurement. To narrow the resonance peaks, the microwave power for the controlled rotation is decreased by 12 dB from the values used for single-qubit rotations. vBi (i = 2, 3) represents a virtual barrier gate voltage. b,c, Results of the exchange spectroscopy measurements. In each figure, the separation of the two peaks corresponds to the exchange coupling. The background slope of the resonance frequency is due to the displacement of the quantum dot position in the micro-magnet field gradient. The frequency offset from the values in Methods is due to the decay of the persistent current in the superconducting magnet. d, Schematic sequence of the residual exchange-coupling measurement. e,f, Results of the measurement of residual exchange couplings between neighbouring qubits. Each dataset is fit with a sinusoidal function P(tevol) = Vsin(πtevolJoff) to extract the residual exchange coupling Joff. V is the visibility of the oscillation. The errors are 1σ from the mean. g, Schematic sequence of the decoupled CZ oscillation measurement. h,i Typical decoupled CZ oscillations. The solid lines show the fit to a Gaussian decay. The decay times are 3.27 ± 0.08 μs (h) and 5.2 ± 0.3 μs (i). Here we adjust the virtual barrier gate voltages so that the exchange coupling is roughly 10 MHz. All errors are 1σ from the mean.
