Figure 1

The simulated eigenfrequency vs control flux for qubit $A$, described below, for $\Phi ={\Phi }_0$. The inset is a schematic representation of the bare qubit coupled to a harmonic oscillator. The mutual inductance ($M$ in Fig. 3) between the bare qubit and the transmission line is 200 pH.Reuse & Permissions

Figure 2

(a) Simulated current in the pickup loop vs flux $\Phi$, parametrized by the control flux ${\Phi }_c$. (b) The corresponding experimental data, measured for our qubit $A$ (described below) at a flux sweep frequency of 13 Hz. The square symbol marks the measurement point, the open dot the operating point, and the triangle the reset point. The solid vertical line is the “$S$ line” or symmetric-well line. The solid horizontal line near the operating point represents the range of flux biases used to collect the data shown in Fig. 4b. The measured hysteresis is smaller than the simulated because the simulation does not include thermal or quantum fluctuations. The tilting of the measured data is a result of inadvertent mutual inductance between the control-flux drive loop and the flux loop. The hysteresis seen in the experimental data near $\Phi =0.7{\Phi }_0$ is hysteresis of the measurement circuit and not of the qubit.Reuse & Permissions

Figure 3

Schematic of the experimental setup.Reuse & Permissions

Figure 4

(a)–(c) Results for qubit $A$. (a) The probability of switching from the initial starting well (left) to the other well (right) after application of a control-flux pulse vs pulse duration for ${\Phi }_c=0.6{\Phi }_0$. Each data point represents the average of 460 single shot measurements. (b) The probability of switching vs pulse duration for seven uniformly spaced values of flux ($\pm 0.035{\Phi }_0$) at ${\Phi }_c=0.43{\Phi }_0$, as depicted by the horizontal lines of length $0.07{\Phi }_0$ in Fig. 2. (c) shows the probability of switching as the pulse duration is made longer. The observed Larmor oscillation frequency of the upper three panels is 1.28 GHz. (d) plots the Larmor oscillations of qubit $B$ using only the arbitrary waveform generator as a pulse source. The oscillation frequency is 1.44 GHz and the coherence time is substantially longer than 35 ns. The poorly defined behavior below 16 ns and the gap in the oscillations from 16 to 21 ns result from pulse distortions produced by the arbitrary waveform generator. When the pulse length is less than a few times the inverse of the maximum clock frequency (200 MHz), the resulting pulse waveform is severely distorted.Reuse & Permissions