Bulletin of the American Physical Society
APS March Meeting 2022
Volume 67, Number 3
Monday–Friday, March 14–18, 2022; Chicago
Session W40: Noise Reduction and Error Mitigation in Quantum Computing IIFocus Recordings Available
|
Hide Abstracts |
Sponsoring Units: DQI Chair: Youngseok Kim, IBM Room: McCormick Place W-196B |
Thursday, March 17, 2022 3:00PM - 3:12PM |
W40.00001: Efficient Analysis of Low Frequency Noise in Quantum Devices Tommy O Boykin II, Michael D Stewart, Antonio L Levy, Felix F Borjans, Roman Caudillo, Florian Luthi, Jim S Clarke Noise can limit the prospects of semiconducting quantum devices through its effect on the coherence time, gate fidelities, readout fidelities, and device integration. Improving these prospects first requires characterizing the noise and the variations within a device and between devices of different design. Typically, noise is analyzed in the frequency domain via the power spectral density (PSD) and parameterized by a coefficient at 1 Hz and a single power law exponent, S = Afα. Often, however, data are not well modeled by a single power law and hint at a stronger frequency dependence below ∼ 1 Hz. Thus, to obtain representative behavior, many time-consuming measurements must be taken creating a bottleneck in device feedback. Here, we present systematic measurements of the noise in industrially fabricated metal-oxide-semiconductor quantum dots analyzed in the frequency and time domains using a technique borrowed from the atomic clock community known as the Allan variance. The combination of these techniques allows for more complete characterization of the noise than is possible with the PSD alone. It also provides a map onto integer noise exponents (such as α = -1 and α = -2) and, with modification, a measure of the expected drift in the device as a function of time. |
Thursday, March 17, 2022 3:12PM - 3:24PM |
W40.00002: SPAM-Robust Quantum Noise Spectroscopy Muhammad Q Khan, Leigh M Norris, Wenzheng Dong, Lorenza Viola, Muhammad Q Khan Existing quantum noise spectroscopy (QNS) protocols allow for the reconstruction of the power spectral density of the noise experienced by a qubit. However, they are limited by the assumption that no state preparation and measurement (SPAM) errors are present, which is invalid in many of the current NISQ-era devices. We propose a modification to existing spin-locked based protocols that make them largely insensitive to the SPAM errors. Our modified protocol not only improves the reconstruction of the target noise spectra but, for noise sources that are non-classical, it also allows for the quantification of state preparation and measurement errors separately. |
Thursday, March 17, 2022 3:24PM - 3:36PM |
W40.00003: Evanescent-wave Johnson noise from BCS superconductors Hruday D Mallubhotla, Robert J Joynt, Maxim G Vavilov Qubits near conducting devices are susceptible to decoherence due to electromagnetic field fluctuations, which leak from the devices as evanescent-wave Johnson noise. An interesting question is to what extent these fluctuations change when the metal becomes superconducting. This noise depends on the electromagnetic response function, which for BCS superconductors has been well studied for arbitrary impurity density. We use this response function to calculate the noise outside the surface of a superconducting half-space in thermodynamic equilibrium. We present T1 of a qubit as a function of temperature, qubit frequency, and distance from the surface. This enables us to characterize the transition from the normal to the superconducting state and we find that the surface-wave contribution is greater in the superconductor. In many experiments a superconducting device element is out of equilibrium due to quasiparticle poisoning, and we also investigate this case. The results show how to use charge or spin qubits as probes of superconducting devices, and they can serve as guides in the design of qubit device architectures. |
Thursday, March 17, 2022 3:36PM - 3:48PM |
W40.00004: Non-Markovian Noise as a Resource for Suppressing Markovian Errors in Superconducting Qubits Evangelos Vlachos, Jeffrey S Marshall, Haimeng Zhang, Tameem Albash, Eli Levenson-Falk, Vivek Maurya Non-Markovian noise environments, i.e. environments with temporal correlation, are well known to have detrimental effects on quantum computational tasks with superconducting qubits. In recent years, there has been intensive interest on characterizing such environments, and on active control protocols to eliminate their effects. In this talk, we present computational and experimental results showing that non-Markovian noise can indeed be used to improve the coherence of a qubit embedded in a purely Markovian noisy background. We further show that our quantum trajectory simulations enable us to find the memory kernel function that offers the best improvement in qubit coherence. We compare these computational/experimental results with theoretical predictions of the corresponding master equation and show that this stochastic error correction scheme yields even better performance than predicted by theory. We finally discuss how this method compares with conventional error correcting schemes and how our results provide a powerful tool in controlling and engineering qubit dissipation processes. |
Thursday, March 17, 2022 3:48PM - 4:00PM |
W40.00005: Quantum Markov Semigroups with Hamiltonian Terms and Zeno Effect Nicholas D LaRacuente Quantum Markov semigroups (QMSs) model time-evolution of open quantum systems in dissipative environments. It is known that any finite-dimensional, continuous QMS with a particular detailed balance condition induces exponential decay of a state's relative entropy to its fixed point projection, following a modified logarithmic-Sobolev inequality (MLSI). In contrast, questions have lingered about semigroups that include a non-trivial Hamiltonian term (precluding detailed balance) and discrete time analogs. We generalize some ideas from semigroups to discrete processes that may include rotations. We then show counter-examples to MLSI and barriers to decay for continuous semigroups with Hamiltonians. When a Hamiltonian shrinks the overall fixed point subspace, strong noise induces a generalized Zeno effect. Counter-intuitively, decay to the long-time fixed point sometimes slows with increasing noise strength, as the Zeno and long-time limits differ. We analytically lower bound rates of convergence to long-time and Zeno fixed points. We numerically analyze simple examples. Finally, we present an experiment run on an IBM Quantum device, in which frequently depolarizing one qubit protects an interacting neighbor from dephasing. |
Thursday, March 17, 2022 4:00PM - 4:12PM |
W40.00006: Dynamical decoupling of two-level systems by quantum Zeno effect Xinyuan You, Ziwen Huang, Shaojiang Zhu, Anna Grassellino, Alexander Romanenko The lifetime of superconducting qubits is currently limited by losses from tunneling two-level systems (TLSs) in the amorphous oxide layers. Recent experiments indicate that qubit decay is dominated by several TLSs that are near-resonant with the qubit. In this presentation, we propose to decouple those TLSs from the qubit by exploiting the quantum Zeno effect. We provide numerical evidence of improved qubit lifetime, and comment on possible experimental realizations. |
Thursday, March 17, 2022 4:12PM - 4:48PM |
W40.00007: Linking many-body physics to many-time physics: Characterising micro and macro features of non-Markovian quantum processes Invited Speaker: Kavan Modi A classical stochastic process is a joint probability distribution of a random variable over time. Its quantum generalisation then turns out to be a multipartite density matrix. We refer to the studies of this density matrix as many-time physics in analogy to the well-founded field of many-body physics. Here, we report a set of tools that allows us to characterise both the detailed features of a quantum process, as well as its coarse structures. The former, we show, could be used to tame correlated noise due to a quantum stochastic process. The latter, on the other hand, allows us to explore exotic features such as genuine multipartite entanglement in time. All of these tools are well-tested and shown to be highly effective on NISQ devices with real noise. Importantly, these tools have direct application for noise reduction for NISQ devices and studying facets of complex quantum processes. |
Thursday, March 17, 2022 4:48PM - 5:00PM |
W40.00008: Measuring the Stability of NISQ Gate-Based Hardware Using a 1+1 Quantum Field Theory Patrick Dreher, Kubra Yeter Aydeniz, Alexander F Kemper, Raphael Pooser, Zachary Parks, Erik Gustafson, Yannick L Meurice, Aadithya Nair Coherent and incoherent errors present on today's Noisy Intermediate Scale Quantum (NISQ) processors impact the quantum computational accuracy and reproducibility of physics applications run on these machines. Understanding and quantitatively measuring these errors provide essential information that assists in interpreting the results of these quantum computations. We report here on an in-depth study using cycle benchmarking to measure the impact of these errors on the stability of the 2-qubit CNOT gates in the 1+1 Transverse Field Ising Model (TFIM) circuit. Measurements of inter-day and intra-day qubit calibration drift and placement of the quantum circuit on separate qubit groups in different physical locations on the processor are presented using this TFIM Hamiltonian running on an IBM Quantum Network superconducting transmon hardware platform. Studies are also getting underway that examine magnon spectra and scattering phase shift quantum computations. All of these results are summarized in the larger context of their impacts on physics applications implemented on NISQ type gate based quantum computing hardware platforms. |
Thursday, March 17, 2022 5:00PM - 5:12PM |
W40.00009: Decoherence under two perpendicular noises Guy Ramon, Lukasz Cywinski Many solid-state qubit systems are afflicted by noise mechanisms that operate along two perpendicular axes of the Bloch sphere. Depending on the qubit's control fields, either noise can be longitudinal or transverse to the qubit's quantization axis, thus affecting its dynamics in distinct ways, generally contributing to decoherence that goes beyond pure dephasing. Here we present a theory that provides a unified platform to study dynamics of a qubit subjected to two perpendicular noises under dynamical decoupling pulse sequences. The theory is demonstrated by the commonly encountered case of power-law noise spectra. |
Thursday, March 17, 2022 5:12PM - 5:24PM |
W40.00010: Correlated many-body noise and emergent 1/f behavior Thomas P Lloyd, Hossein R Sadeghpour, Valentin Walther Fluctuating electric fields emanating from surfaces are a major possible source of decoherence in a number of quantum applications, including trapped ions and near-surface NV diamond qubits. We show that at low temperatures, due to superradiant decay, phonon-induced excitation exchange between adsorbed atoms can counterintuitively mitigate the electric field noise. This contrasts with perhaps the anticipated behavior of corrrelated dynamics amplifying detrimental noise. We derive an exact mapping between the noise spectrum of N interacting fluctuators with M vibrational levels to (N+M-1 \choose N)-1 noninteracting two-level dipoles. The anharmonic interaction of the fluctuators with the surface is semiempirical and physically motivated. This anharmonicity affects the noise spectral power intensity at higher temperatures which we simulate numerically. We describe conditions for which the ubiquitous 1/f noise emerges naturally from the coupled dynamics of, remarkbly, identical fluctuators and whose behavior depends critically on correlation among the fluctuators. We believe this work constitutes the first derivation of correlated superradiant noise and emergent 1/f behavior. |
Thursday, March 17, 2022 5:24PM - 5:36PM |
W40.00011: Large fluctuations of qubit decoherence by 1/f noise of a bath of two-level fluctuators Mohammad Mehmandoost, Viatcheslav V Dobrovitski Decoherence by the 1/f noise is a serious problem for superconductor and quantum dot-based quantum computing platforms. The 1/f noise is produced by a bath of Two-Level Fluctuators (TLFs), each TLF randomly changing its state at a rate γ; an ensemble of TLFs with a logarithmically uniform distribution of rates produces 1/f noise [1]. Within Gaussian approximation, the qubit decoherence is determined by the noise first spectral density [2]. However, the validity region of this approximation is poorly understood, and little is known about decoherence outside of this regime [3]. |
Thursday, March 17, 2022 5:36PM - 5:48PM |
W40.00012: Universal fidelity reduction of quantum operations from weak dissipation Tahereh Abad, Göran Johansson, Anton F Kockum, Jorge Fernández Pendás Quantum information processing is in real systems often limited by dissipation, stemming from remaining uncontrolled interaction with microscopic degrees of freedom. Given recent experimental progress, we consider weak dissipation, resulting in a small error probability per operation. Here, we find a simple formula for the fidelity reduction of any desired quantum operation and interestingly this reduction is independent of the specific gate and depends only on the operation time and the dissipation. Using this formula, we investigate the situation where dissipation in different parts of the system have correlations, which is detrimental for the successful application of quantum error correction. Surprisingly, we find that a large class of correlations gives the same fidelity reduction as uncorrelated dissipation of similar strength. |
Thursday, March 17, 2022 5:48PM - 6:00PM |
W40.00013: Distinguishing between quantum and classical Markovian dephasing dissipation Alireza Seif, Yuxin Wang, Aashish Clerk Understanding whether dissipation in an open quantum system is truly quantum is a question of both fundamental and practical interest. We consider a general model of n qubits subject to correlated Markovian dephasing, and present a sufficient condition for when bath-induced dissipation can generate system entanglement and hence must be considered quantum. Surprisingly, we find that the presence or absence of time-reversal symmetry (TRS) plays a crucial role: broken TRS is required for dissipative entanglement generation. Further, simply having non-zero bath susceptibilities is not enough for the dissipation to be quantum. Our work also presents an explicit experimental protocol for identifying truly quantum dephasing dissipation, and lays the groundwork for studying more complex dissipative systems and finding optimal noise mitigating strategies. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700