Quantum amplifiers are intrinsically nonlinear systems whose performance limits are set by quantum mechanics. In quantum measurement, amplifier operation is conventionally optimized in the linear regime by maximizing signal-to-noise ratio, an objective that is well-suited to parameter estimation but is typically insufficient for more general tasks such as arbitrary quantum state discrimination. Here we show that single-shot quantum state classification can benefit from operating a quantum amplifier outside the linear regime, when the measurement chain is optimized end-to-end for a task-specific cost function. We analyze a realistic superconducting readout architecture that includes state preparation, cryogenic nonlinear amplification, and room-temperature detection with finite noise. By introducing performance metrics tailored to state discrimination, we identify operating regimes in which nonlinear amplification provides a measurable advantage and clarify the trade-offs that ultimately limit classification fidelity. Building on these results, we propose a qubit readout architecture without cavity displacement that exploits nonlinear amplification to enhance single-shot state discrimination performance. Our results establish the practical value of nonlinear quantum amplifiers for quantum state discrimination and lay the foundation for a broader program to develop a general, end-to-end framework for resource-constrained optimization of nonlinear amplification in quantum information processing tasks.
Comments
Added a new section in version 2, corrected grant number in version 3
Quantum amplifiers are intrinsically nonlinear systems whose performance limits are set by quantum mechanics. In quantum measurement, amplifier operation is conventionally optimized in the linear regime by maximizing signal-to-noise ratio, an objective that is well-suited to parameter estimation but is typically insufficient for more general tasks such as arbitrary quantum state discrimination. Here we show that single-shot quantum state classification can benefit from operating a quantum amplifier outside the linear regime, when the measurement chain is optimized end-to-end for a task-specific cost function. We analyze a realistic superconducting readout architecture that includes state preparation, cryogenic nonlinear amplification, and room-temperature detection with finite noise. By introducing performance metrics tailored to state discrimination, we identify operating regimes in which nonlinear amplification provides a measurable advantage and clarify the trade-offs that ultimately limit classification fidelity. Building on these results, we propose a qubit readout architecture without cavity displacement that exploits nonlinear amplification to enhance single-shot state discrimination performance. Our results establish the practical value of nonlinear quantum amplifiers for quantum state discrimination and lay the foundation for a broader program to develop a general, end-to-end framework for resource-constrained optimization of nonlinear amplification in quantum information processing tasks.
Although linear quantum amplification has proven essential to the processing of weak quantum signals, extracting higher-order quantum features such as correlations in principle demands nonlinear operations. However, nonlinear processing of quantum signals is often associated with non-idealities and excess noise, and absent a general framework to harness nonlinearity, such regimes are typically avoided. Here we present a framework to uncover general quantum signal processing principles of a broad class of bosonic quantum nonlinear processors (QNPs), inspired by a remarkably analogous paradigm in nature: the processing of environmental stimuli by nonlinear, noisy neural ensembles, to enable perception. Using a quantum-coherent description of a QNP monitoring a quantum signal source, we show that quantum nonlinearity can be harnessed to calculate higher-order features of an incident quantum signal, concentrating them into linearly-measurable observables, a transduction not possible using linear amplifiers. Secondly, QNPs provide coherent nonlinear control over quantum fluctuations including their own added noise, enabling noise suppression in an observable without suppressing transduced information, a paradigm that bears striking similarities to optimal neural codings that allow perception even under highly stochastic neural dynamics. Unlike the neural case, we show that QNP-engineered noise distributions can exhibit non-classical correlations, providing a new means to harness resources such as entanglement. Finally, we show that even simple QNPs in realistic measurement chains can provide enhancements of signal-to-noise ratio for practical tasks such as quantum state discrimination. Our work provides pathways to utilize nonlinear quantum systems as general computation devices, and enables a new paradigm for nonlinear quantum information processing.
Linear quantum amplifiers are indispensable tools for quantum technologies, yet their performance is fundamentally limited by quantum noise, precluding any signal-to-noise ratio (SNR) enhancement unless supplemented by post-selection or non-classical resources. To surpass this limitation, we propose a nonlinear quantum amplification strategy that exploits the interplay between a gain-stabilized bright eigenmode of a coupled two-mode bosonic system and Kerr nonlinearity. We demonstrate that this interplay enables the signal gain to surpass the noise gain in a selected quadrature, leading to a net increase in the SNR beyond the quantum limits of conventional linear amplifiers. Our work thus establishes a novel nonlinear amplification paradigm capable of enhancing the SNR, with promising applications across quantum information processing, quantum communications, and quantum metrology.
Efficient quantum state measurement is important for maximizing the extracted information from a quantum system.For multiqubit quantum processors, in particular, the development of a scalable architecture for rapid and high-fidelity readout remains a critical unresolved problem.Here we propose reservoir computing as a resource-efficient solution to quantum measurement of superconducting multiqubit systems.We consider a small network of Josephson parametric oscillators, which can be implemented with minimal device overhead and in the same platform as the measured quantum system.We theoretically analyze the operation of such a device as a reservoir computer to classify stochastic timedependent signals subject to quantum statistical features.We apply this reservoir computer to the task of multinomial classification of measurement trajectories from joint multiqubit readout.For a 2-qubit dispersive measurement under realistic conditions we demonstrate a classification fidelity reliably exceeding that of an optimal linear filter using only 2-5 reservoir nodes, while simultaneously requiring far less calibration data-as little as a few shots per state.We understand this remarkable performance through an analysis of the network dynamics and develop an intuitive picture of reservoir processing generally.Finally, we demonstrate how to operate this device to perform 2-qubit state tomography and continuous parity monitoring with equal effectiveness and ease of calibration.This reservoir processor avoids computationally intensive training common to other machine learning frameworks and can be implemented as an integrated cryogenic superconducting device for low-latency processing of quantum signals on the computational edge.
A quantum-limited amplifier enables the amplification of weak signals while introducing minimal noise dictated by the principles of quantum mechanics. Such amplifiers serve a broad spectrum of applications in quantum computing, including fast and accurate readout of superconducting qubits and spins, as well as various uses in quantum sensing and metrology. Parametric amplification, primarily developed with use of Josephson junctions, has evolved into the leading technology for highly effective microwave measurements within quantum circuits. Despite their significant contributions, these amplifiers face fundamental limitations, such as their inability to handle high powers, their sensitivity to parasitic magnetic fields, and particularly their limitation to operate only at millikelvin temperatures. To tackle these challenges, here we experimentally develop a novel quantum-limited amplifier based on superconducting kinetic inductance and present an extensive theoretical model to describe this nonlinear coupled-mode system. Our device surpasses the conventional constraints associated with Josephson-junction amplifiers by operating at much higher temperatures up to 4.5 K. With two distinct spectral modes and tunability through bias current, this amplifier can operate selectively in both the single-mode-amplification regime and the double-mode-amplification regime near the quantum noise limit. Using a nonlinear thin film exhibiting kinetic inductance, our device attains gain exceeding 50 dB in a single-mode configuration and 32 dB in a double-mode configuration, while adding 0.82 input-referred noise quanta. Importantly, this amplifier eliminates the need for Josephson junctions, resulting in the capability to handle significantly higher powers than Josephson junction--based amplifiers. It also demonstrates resilience in the presence of magnetic fields, offers a straightforward design, and increases reliability. This positions the amplifier as a versatile solution for quantum applications and facilitates its integration into future superconducting quantum computers.
Superconducting parametric amplifiers (SPAs) are critical components for ultralow-noise qubit readout in quantum computing, addressing the critical challenge of amplifying weak quantum signals without introducing noise that degrades coherence and computational fidelity. Unlike classical amplifiers, SPAs can achieve or closely approach quantum-limited performance, specifically the Standard Quantum Limit (SQL) of half a photon of added noise for phase-preserving amplification. The core principle of SPAs relies on parametric amplification, where energy is transferred from a strong pump tone to a weak input signal through non-dissipative nonlinear mixing processes. This is enabled by intrinsic nonlinearities in superconducting materials, primarily kinetic inductance in thin films (e.g., NbTiN, Al) and, more significantly, the Josephson effect in Josephson junctions. These nonlinear elements facilitate frequency mixing (three-wave or four-wave mixing) and can operate in phase-preserving or phase-sensitive amplification modes, with the latter allowing for noise squeezing below the SQL. This chapter emphasizes the significant role of resonator design in determining critical SPA performance metrics such as gain, bandwidth, and noise characteristics. It details both lumped-element (LC) and distributed-element (coplanar waveguide, CPW) resonators, discussing their unique properties, suitability for different frequency ranges, and the importance of achieving high-quality factors (Q) for efficient energy storage and minimal loss. A practical design and simulation of a meandered quarterwavelength CPW resonator coupled to a feed line is presented, illustrating how precise control over geometric parameters optimizes resonant frequency, coupling strength, and quality factor for high-fidelity qubit state discrimination.
Single-shot dispersive readout of superconducting qubits is a very important part of real-time quantum feedback and quantum error correction (QEC). Conventional discriminators such as thresholding and matched filtering are optimal under Gaussian noise assumptions, yet performance degrades in the presence of non-Gaussian effects in particular when including energy relaxation during readout and leakage to non-computational states. Neural network based discriminators have been proposed to mitigate these limitations. There is still a lack of characterization of the scenarios in which they offer a significant benefit and the mechanisms that underlie that advantage. This paper shows a controlled simulation study of six discrimination methods that get more detailed as the model gets more complex: thresholding, matched filtering, linear discriminant analysis, a multilayer perceptron operating on integrated in-phase and quadrature (IQ) features, one-dimensional convolutional neural network operating on the full time-resolved measurement trace, and a hybrid CNN. By systematically varying physically relevant parameters, we map classifier performance as a function of the ratio of readout duration to qubit relaxation time, which governs the strength of non-Gaussian effects. We demonstrate that neural network-based discriminators offer minimal or no benefits in Gaussian-dominated contexts where linear approaches are almost optimal, yet produce substantial enhancements when non-Gaussian effects are present. Specifically, a hybrid architecture combining a small temporal CNN branch with a direct skip connection for the integrated IQ features achieves assignment fidelity of about 0.98 in the stress-test scenario, which represents a significant improvement over the matched filter. This hybrid architecture, at only 686 trainable parameters, is smaller than the pure CNN yet dominates it in easy settings (where the pure CNN underperforms linear baselines by ≈6 percentage points, pp) while retaining its advantage in hard settings. Additionally, confusion-matrix analysis reveals that this enhancement results from the capacity of temporal models to leverage within-window dynamics that are unavailable to classifiers utilizing integrated observations. Rather than recommending the hybrid as a universally superior method, we characterise the regime in which it helps: it trails the matched filter by only ≈1.2 pp in Gaussian conditions while providing a + 5.39 pp advantage in the stress-test regime, which makes it a regime-robust default wherever the operating point is uncertain or leakage is a meaningful error channel. These results provide a predictive understanding of when neural-network-based readout is beneficial and offer practical guidance for selecting discrimination strategies in superconducting qubit systems subject to real-time constraints.