High-frequency readout free from transmon multi-excitation resonances
Pavel D. Kurilovich, Thomas Connolly, Charlotte G. L. Bøttcher, Daniel K. Weiss, Sumeru Hazra, Vidul R. Joshi, Andy Z. Ding, Heekun Nho, Spencer Diamond, Vladislav D. Kurilovich, Wei Dai, Valla Fatemi, Luigi Frunzio, Leonid I. Glazman, Michel H. Devoret · 2025
Read High-frequency readout free from transmon multi-excitation resonances by Pavel D. Kurilovich, Thomas Connolly, Charlotte G. L. Bøttcher, Daniel K. Weiss, Sumeru Hazra, Vidul R. Joshi, Andy Z. Ding, Heekun Nho, Spencer Diamond, Vladislav D. Kurilovich, Wei Dai, Valla Fatemi, Luigi Frunzio, Leonid I. Glazman, Michel H. Devoret (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-09161.txt.
Introduction
Pavel Kurilovich and co-authors present a 2025 paper on reading out superconducting qubits. Quantum error correction needs operations far above accuracy thresholds. In superconducting machines, reading the qubit state is the weakest step. The standard readout tone sits near the qubit's own frequency. At high power, stray resonances in the qubit spectrum fire and spoil the reading. Those multi-excitation resonances cap how strong the tone can be.
The team demonstrates a high-frequency readout that dodges those resonances. A stronger tone can then be used without waking the stray states. Stronger tones read faster and more faithfully. Quantum hardware teams get a path to the accuracy that error correction demands. Measurement is run constantly in error correction, so small gains compound. The work targets the transmon, the field's workhorse qubit, which gives it wide reach.
Catalog record
| CREATOR | Pavel D. Kurilovich, Thomas Connolly, Charlotte G. L. Bøttcher, Daniel K. Weiss, Sumeru Hazra, Vidul R. Joshi, Andy Z. Ding, Heekun Nho, Spencer Diamond, Vladislav D. Kurilovich, Wei Dai, Valla Fatemi, Luigi Frunzio, Leonid I. Glazman, Michel H. Devoret |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 86 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-09161
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
High-frequency readout free from transmon multi-excitation resonances Pavel D. Kurilovich, Thomas Connolly, Charlotte G. L. Bøttcher, Daniel K. Weiss, Sumeru Hazra, Vidul R. Joshi, Andy Z. Ding, Heekun Nho, Spencer Diamond, Vladislav D. Kurilovich, Wei Dai, Valla Fatemi, Luigi Frunzio, Leonid I. Glazman, Michel H. Devoret Abstract Quantum computation will rely on quantum error correction to counteract decoherence. Successfully implementing an error correction protocol requires the fidelity of qubit operations to be well-above error correction thresholds [ 1 , 2 ] . In superconducting quantum computers, measurement of the qubit state remains the lowest-fidelity operation [ 3 ] . For the transmon, a prototypical superconducting qubit, measurement is carried out by scattering a microwave tone off the qubit. Conventionally, the frequency of this tone is of the same order as the transmon frequency. The measurement fidelity in this approach is limited by multi-excitation resonances in the transmon spectrum which are activated at high readout power [ 29 , 5 , 22 , 23 , 8 , 24 , 10 ] . These resonances excite the qubit outside of the computational basis, violating the desired quantum non-demolition character of the measurement. Here, we find that strongly detuning the readout frequency from that of the transmon exponentially suppresses the strength of spurious multi-excitation resonances. By increasing the readout frequency up to twelve times the transmon frequency, we achieve a quantum non-demolition measurement fidelity of 99.93% with a residual probability of leakage to non-computational states of only 0.02%. I Introduction Figure 1: (a) Readout of a transmon (red) is achieved by monitoring the signal reflected from the readout resonator (blue). The information about the transmon state is encoded in the phase of the reflected signal. (b) In our experiment, the frequency of the readout resonator exceeds that of the transmon roughly by a factor of 12. (c) Transition spectrum of the system in the standard readout regime, \omega_{\mathrm{r}}\sim\omega_{\mathrm{q}} . Red lines show transition frequencies from the ground state |0\rangle of the transmon as a function of the offset charge n_{g} . The shaded region indicates transitions to states within the Josephson energy well. Blue lines correspond to n -photon transitions of the readout resonator. The number of accidental spectral collisions between the transmon and the readout resonator is large. These collisions are responsible for the undesired transmon excitation to non-computational states. (c) In the regime \omega_{\mathrm{r}}\gg\omega_{\mathrm{q}} , the number of spectral collisions is greatly reduced. Inset: the resonator transition anti-crosses with a transmon transition to a highly lying state |j\rangle . The strength \Omega of this anti-crossing is proportional to the matrix element of the transmon charge operator N . (d) Numerically calculated matrix element \langle 0|N|j\rangle for different final states |j\rangle as a function of the transition frequency \omega=E_{j0}/\hbar . The matrix element is exponentially suppressed with increasing \omega/\omega_{\mathrm{q}} . The blue line corresponds to the resonator frequency. Figure 2: Numerically simulated landscape of multi-excitation resonances for a transmon with E_{J}/E_{C}=60 at offset charge n_{g}=0.25 . The horizontal axis shows the normalized drive frequency and the vertical axis shows the normalized drive power (quantified by the AC Stark shift \delta\omega of the qubit transition). The color shows the degree of hybridization \Theta between the computational state |1\rangle and the non-computational states; multi-excitation resonances correspond to the regions where this hybridization is enhanced. Specifically, dimensionless number \Theta corresponds to the probability of finding the system outside of the computational subspace due to accidental Rabi processes involving the non-computational states. Away from multi-excitation resonances \Theta=0 , on any of the isolated resonances \Theta=1/2 . (a) Multi-excitation resonances in a broad frequency range. The blue star shows the frequency and the average drive power of our chosen readout pulse. (b)-(d) Zoom in on three frequency windows highlighted with colored boxes in panel (a). (b) Standard regime of the transmon readout, \omega\sim\omega_{\mathrm{q}} . Multi-excitation resonances quickly become dense with the increase of power. (c,d) As the frequency is increased, both the number of the multi-excitation resonances and the corresponding transition amplitudes are progressively suppressed. Many quantum error correction protocols, such as the surface code [ 11 ] , check for errors by repeatedly measuring a subset of qubits [ 12 , 13 , 14 , 15 , 16 , 17 , 3 ] . These error checks should not introduce additional errors, which imposes two stringent requirements on the measurement. First, the measurement must project the system onto an eigenstate of the observed quantity. Second, the eigenstate must be successfully identified. A measurement that satisfies these criteria is called quantum non-demolition (QND) [ 18 ] . Achieving a high-fidelity QND qubit measurement remains one of the outstanding challenges in performing real-time quantum error correction across different quantum computing platforms [ 3 , 19 , 20 , 21 ] . In circuit quantum electrodynamics, qubit measurement is carried out by scattering non-resonant microwaves off the qubit and analyzing the amplified [ 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 ] scattered signal [ 9 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 20 ] . Information about the state of the qubit is encoded in the amplitude and phase of the scattered signal, see Fig. 1 (a). Since the readout tone in such a dispersive readout is not resonant with the qubit, the measurement should not cause transitions between energy eigenstates. Thus, dispersive measurement can in principle be QND. One limitation on the QND character of the readout comes from the possibility of qubit decay during the measurement [ 41 ] . This limitation can be overcome by increasing the measurement speed, which requires increasing the power of the readout tone. However, there is experimental evidence that dispersive readout loses its QND character at high powers [ 35 , 29 , 37 , 5 , 39 , 17 ] . Transmons [ 12 , 44 ] , the most commonly used superconducting qubits, get excited to higher non-computational states by strong readout tones. It is believed that this happens due to the non-linear absorption of readout photons [ 22 , 23 , 8 , 24 , 10 ] . In this process, a few photons from the readout tone are simultaneously absorbed by the transmon, causing it to “leak” to a non-computational state. Nonlinear absorption can be resonant even when the readout frequency is detuned from the qubit transition frequency. Problematically, the strength of multi-excitation resonances, and thus the leakage probability, sharply increases with the power of the readout tone. Error correction protocols are designed to correct only for bit-flip and phase-flip errors, not leakage errors [ 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 ] . Multi-excitation resonances thus must be eliminated for quantum computation to operate. Is it possible to avoid the activation of transmon multi-excitation resonances and implement high-fidelity QND qubit readout? In this article, we answer this question affirmatively by demonstrating a dispersive transmon readout free of significant multi-excitation resonances. We achieve this by using a readout resonator with a frequency \omega_{\mathrm{r}} exceeding that of the transmon, \omega_{\mathrm{q}} , by a factor of 12, see Fig.