Classical artificial neural networks have witnessed widespread successes in machine-learning applications. Here, we propose fermion neural networks (FNNs) whose physical properties, such as local density of states or conditional conductance, serve as outputs, once the inputs are incorporated as an initial layer. Comparable to back propagation, we establish an efficient optimization, which entitles FNNs to competitive performance on challenging machine-learning benchmarks. FNNs also directly apply to quantum systems, including hard ones with interactions, and offer in situ analysis without preprocessing or presumption. Following machine learning, FNNs precisely determine topological phases and emergent charge orders. Their quantum nature also brings various advantages: quantum correlation entitles more general network connectivity and insight into the vanishing gradient problem, quantum entanglement opens up alternative avenues for interpretable machine learning, etc.

Quantum computers based on gate-defined quantum dots (QDs) are expected to scale. However, as the number of qubits increases, the burden of manually calibrating these systems becomes unreasonable and autonomous tuning must be used. There has been a range of recent demonstrations of automated tuning of various QD parameters such as coarse gate ranges, global state topology (e.g., single QD, double QD), charge, and tunnel coupling with a variety of methods. Here, we demonstrate an intuitive, reliable, and data-efficient set of tools for an automated global state and charge tuning in a framework deemed physics-informed tuning (PIT). The first module of PIT is an action-based algorithm that combines a machine learning classifier with physics knowledge to navigate to a target global state. The second module uses a series of one-dimensional measurements to tune to a target charge state by first emptying the QDs of charge, followed by calibrating capacitive couplings and navigating to the target charge state. The success rate for the action-based tuning consistently surpasses $95\mathrm{\%}$ on both simulated and experimental data suitable for off-line testing. The success rate for charge setting is comparable when testing with simulated data, at $95.5(5.4)\mathrm{\%}$, and only slightly worse for off-line experimental tests, with an average of $89.7(17.4)\mathrm{\%}$ (median $97.5\mathrm{\%}$). It is noteworthy that the high performance is demonstrated both on data from samples fabricated in an academic cleanroom as well as on an industrial 300-mm process line, further underlining the device agnosticism of PIT. Together, these tests on a range of simulated and experimental devices demonstrate the effectiveness and robustness of PIT.

Future large-scale quantum computing devices based on semiconductor spin qubits require multiplexed microwave control signals for manipulation and readout of the qubits. However, interfacing these control signals between room temperature and cryogenic temperatures where the qubits operate is a complex technical challenge. Here, we show a microwave modulator based on a nanoengineered semiconductor quantum dot that is designed to operate at millikelvin temperatures alongside the qubits. We operate the modulator as a mixer and a frequency multiplier over a bandwidth of up to 25 GHz. We estimate the power dissipation when driving quantum gates using frequency-up-converted signals under realistic experimental conditions to be 4 pW, highly compatible with the cooling powers available in current commercial dilution refrigerators. The device is fabricated in silicon using atomic precision lithography, providing a pathway toward combining qubits and classical control functionality on the same integrated chip.

The long-lived and optically addressable spin states of silicon vacancies ($V_{\mathrm{Si}}$) in $4H$-SiC make them promising qubits for quantum communication and sensing. These color centers can be created in both the hexagonal ($V1$) and in the cubic ($V2$) local crystallographic environments of the $4H$-SiC host. While the spin of the $V2$ center can be efficiently manipulated by optically detected magnetic resonance at room temperature, spin control of the $V1$ center above cryogenic temperatures has so far remained elusive. Here, we show that the dynamic strain of surface acoustic waves can overcome this limitation and efficiently excite magnetic resonances of $V1$ centers up to room temperature. Based on the width and temperature dependence of the acoustically induced spin resonances of the $V1$ centers, we attribute them to transitions between spin sublevels in the excited state. The acoustic spin control of both kinds of $V_{\mathrm{Si}}$ centers in their excited states opens alternative ways for applications in quantum technologies based on spin optomechanics.

The local oscillator in a practical continuous-variable quantum key distribution system fluctuates at any time during the key distribution process, which may open security loopholes for the eavesdropper to hide her eavesdropping behaviors. Based on this, we investigate a more stealthy quantum attack where the eavesdroppers simulate random fluctuations of local oscillator intensity in a practical discrete-modulated continuous-variable quantum key distribution system. Theoretical simulations show that both communicating parties will misestimate channel parameters and overestimate the secret key rate due to the modified attack model, even though they have monitored the mean local oscillator intensity and shot noise as commonly used. Specifically, the eavesdropper’s manipulation of random fluctuations in LO intensity disturbs the parameter estimation in a realistic discrete-modulated continuous-variable quantum key distribution system, where the experimental parameters are always used for constraints of the semidefinite program modeling. The modified attack introduced by random fluctuations of local oscillator can only be eliminated by monitoring the local oscillator intensity in real time, which places a higher demand on the accuracy of monitoring technology. Moreover, similar quantum hacking will also occur in a practical local local oscillator system by manipulating the random fluctuations in pilot intensity, which shows the strong adaptability and the role of the proposed attack.

The challenge of pattern recognition is to invoke a strategy that can accurately extract features of a dataset and classify its samples. In realistic scenarios, this dataset may be a physical system from which we want to retrieve information, such as in the readout of optical classical memories. The theoretical and experimental development of quantum reading has demonstrated that the readout of optical memories can be significantly enhanced through the use of quantum resources (namely, entangled input states) over that of the best classical strategies. However, the practicality of this quantum advantage hinges upon the scalability of quantum reading, and up to now its experimental demonstration has been limited to individual cells. In this work, we demonstrate quantum advantage at a fixed resource, namely, at fixed mean probe energy, in the multicell problem of pattern recognition. Through experimental realizations of digits from the Modified National Institute of Standards and Technology (MNIST) handwritten digit dataset, and the application of advanced classical postprocessing, we report the use of entangled probe states and photon counting to achieve quantum advantage in classification error over that achieved with classical resources, confirming that the advantage gained through quantum sensors can be sustained throughout pattern recognition and complex postprocessing. This motivates future developments of quantum-enhanced pattern recognition of bosonic loss within complex domains.

Timing requirements for long-range quantum networking are driven by the necessity of synchronizing the arrival of photons, from independent sources, for Bell-state measurements. Thus, characteristics such as repetition rate and pulse duration influence the precision required to enable quantum networking tasks such as teleportation and entanglement swapping. Some solutions have been proposed utilizing classical laser pulses, frequency combs, and biphoton sources. In this article, we explore the utility of the latter method since it is based upon quantum phenomena, which makes it naturally covert, and potentially quantum secure. Furthermore, it can utilize relatively low performance quantum-photon sources and detection equipment, but provides picosecond-level timing precision even under high loss and high noise channel conditions representative of daytime space-Earth links. Therefore, this method is relevant for daytime space-Earth quantum networking and/or providing high-precision secure timing in global positioning system–denied environments.

Classical interferometers are indispensable tools for the precise determination of various physical quantities. Their accuracy is bound by the standard quantum limit. This limit can be overcome by using quantum states or nonlinear quantum elements. Here, we present the experimental study of a nonlinear Josephson interferometer operating in the microwave regime. Our quantum microwave parametric interferometer (QUMPI) is based on superconducting flux-driven Josephson parametric amplifiers combined with linear microwave elements. We perform a systematic analysis of the implemented QUMPI. We find that its interferometric power exceeds the shot-noise limit and observe sub-Poissonian photon statistics in the output modes. Furthermore, we identify a low-gain operation regime of the QUMPI that is essential for optimal quantum measurements in quantum illumination protocols.

Photon loss is the fundamental issue toward the development of quantum networks. To circumvent loss quantum repeaters were proposed, which need high-performance quantum memories. We present an alternative proposal to mitigate photon loss even at large distances and hence to create a global-scale quantum communication architecture without quantum repeaters. In this proposal, photons are sent directly through space, using a chain of co-moving low-earth orbit satellites. This satellite chain would bend the photons to move along the Earth’s curvature and control photon loss due to diffraction by effectively behaving like a set of lenses on an optical table. Numerical modeling of photon propagation through these “satellite lenses” shows that diffraction loss in entanglement distribution can be almost eliminated even at global distances of 20 000 km while considering beam truncation at each satellite and the effect of different errors (e.g., “satellite lens” focal-length fluctuation). In the absence of diffraction loss, the effect of other losses (especially reflection loss) becomes relevant and they are investigated in detail. The total loss is estimated to be less than 30 dB at 20 000 km if other losses are constrained to 2% at each satellite, with 120-km satellite separation and 60-cm diameter satellite telescopes eliminating diffraction loss. Such a low-loss satellite-based optical-relay protocol would enable robust, multimode global quantum communication and would not require either quantum memories or repeater protocol. The protocol can also be the least lossy in almost all distance ranges available (200–20 000 km). Recent advances in space technologies may soon enable an affordable launch facility for such a satellite-relay network. We further introduce the “qubit transmission” protocol, which has a plethora of advantages with both the photon source and the detector remaining on the ground. A specific lens setup was designed for the “qubit transmission” protocol, which performed well in simulation that included atmospheric turbulence in the satellite uplink.

Measuring the losses arising from different materials and interfaces is crucial to improving the coherence of superconducting quantum circuits. Although this has been of interest for a long time, current studies can either only provide bounds to those losses, or require several devices for a complete characterization. In this work, we introduce a method to measure the microwave losses of materials and interfaces with a single multimode superconducting resonator. We demonstrate a formalism for analyzing the loss sensitivity of multimode systems and discuss the design strategies of multimode resonators for material loss studies. We present two types of multimode superconducting resonators for the study of bulk superconductors: the forky whispering-gallery-mode resonator and the ellipsoidal cavity. We use these resonators to measure the surface dielectric, conductor, and seam losses of high-purity (5N5) aluminum and aluminum alloy (6061), as well as how they are affected by chemical etching, diamond turning, and thin-film coating. We find that chemical etching and diamond turning reduce both the surface dielectric and conductive losses of high-purity aluminum, but provide no appreciable improvement to the seam. Coating the surfaces of diamond-turned aluminum alloys with electron-beam-evaporated or sputtered aluminum thin films significantly reduces all three losses under study. In addition, we study the effect of chemical etching on the surface of high-purity aluminum using transmission electron microscopy and find that the chemical etching process creates a thinner and more uniform oxide layer, consistent with the observed improvement in the surface dielectric loss.

In the field of spintronics, there is a strong demand for barrier materials in magnetic tunnel junctions (MTJs) having high tunnel magnetoresistance (TMR) and low resistance area product (RA). However, the design of such barrier materials with atomically controlled composition, vacancies, and disordering of the constituent elements is still challenging due to the combinatorial explosion of potential candidates. Very recently, a quantum annealing (QA) method has been utilized for combinatorial optimization problems in materials science. In this paper, we perform a proof-of-concept study by applying the QA approach combining with first-principles calculations and a machine-learning factorization machine (FM) to MTJs with inverse-type spinel ${\mathrm{Mg}\mathrm{Ga}}_2{\mathrm{O}}_4$ tunnel barrier. We treat 252 combinations of ${\mathrm{Mg}}^{2+}$ and ${\mathrm{Ga}}^{3+}$ cation disordering in the barrier layer of the MTJs and discuss the effect of the cation disordering on the structural stability, TMR, and RA. Our method is superior to simulated annealing, Bayesian optimization, and random sampling in searching for the best MTJs for low total energy and high TMR, but not so for low RA. We also revealed physical origins of high TMR and low RA behind the cation disordering. The present work highlights the applicability and advantage of materials informatics using FM + QA with first-principles calculations in designing spintronic MTJ devices.

Coherent coupling between spin-wave excitations (magnons) and microwave photons in a cavity may disclose new paths to unconventional phenomena as well as novel applications. Here, we present a systematic investigation of yttrium iron garnets (YIG) films on top of coplanar waveguide resonators made of superconducting ${\mathrm{YBa}}_2{\mathrm{Cu}}_3{\mathrm{O}}_7$ (YBCO). We first show that spin-wave excitations with frequency higher than the Kittel mode can be excited by putting in direct contact a 5-$\text{μ}\mathrm{m}$-thick YIG film with the YBCO coplanar resonator (cavity frequency ${\omega }_c/2\pi =8.65$ GHz). With this configuration, we obtain very large values of the collective coupling strength $\lambda /2\pi \approx 2$ GHz and cooperativity $C=5\times {10}^4$. Transmission spectra are analyzed by a modified Hopfield model for which we provide an exact solution that allows us to well reproduce spectra by introducing a limited number of free parameters. It turns out that the coupling of the dominant magnon mode with photons exceeds 0.2 times the cavity frequency, thus demonstrating the achievement of the ultrastrong-coupling regime with this architecture. Our analysis also shows a vanishing contribution of the diamagnetic term, which is a peculiarity of pure spin systems.

We propose a hybrid classical-quantum digitized counterdiabatic algorithm to tackle the protein-folding problem on a tetrahedral lattice. Digitized counterdiabatic quantum computing is a paradigm developed to compress quantum algorithms via the digitization of the counterdiabatic acceleration of a given adiabatic quantum computation. Finding the lowest-energy configuration of the amino-acid sequence is an NP-hard optimization problem that plays a prominent role in chemistry, biology, and drug design. We outperform state-of-the-art quantum algorithms using problem-inspired and hardware-efficient variational quantum circuits. We apply our method to proteins with up to nine amino acids, using up to 17 qubits on quantum hardware. Specifically, we benchmark our quantum algorithm with Quantinuum’s trapped ions, and Google’s and IBM’s superconducting circuits, obtaining high success probabilities with low-depth circuits as required in the noisy intermediate-scale quantum era.

Hybrid quantum systems that utilize controllable coupling between phonons and superconducting qubits could pave the way for a new generation of compact quantum memories and sensor technologies. However, the same electromechanical interaction mechanisms that enable coherent qubit-phonon coupling can also lead to decoherence due to the spontaneous phonon emission. In this paper we study the dynamics of a qubit coupled to an acoustic resonator by a piezoelectric film. By varying the surface topography of the resonator from rough to flat to shaped, we explore the crossover from fast decay of an excited qubit to quantum-coherent coupling between the qubit and an isolated phonon mode. The developed theoretical model allows us to establish and check quantitative links between the results obtained for the rough and shaped topographies. Discrepancy between the theory and the data for the flat resonator points to the imperfections of the studied device. Our experimental approach may be used for precision measurements of crystalline vibrations, the design of quantum memories, and the study of electromechanical contributions to dielectric loss.

The Hong-Ou-Mandel interference between independent photons plays a pivotal role in large-scale quantum networks involving distant nodes. Photons need to work in a pure state for indistinguishability to reach high-quality interference. Also, they need to have a sufficiently long coherence time to reduce the time-synchronization requirements in practical applications. In this paper, we discuss a scheme for generating a pure-state photon-pair source with a long coherence time in periodically poled potassium titanyl phosphate crystals. By selecting the appropriate pump laser and filter, we can simultaneously eliminate the frequency correlation of the parametric photons while achieving a long coherence time. We experimentally develop this pure-state photon-pair source of 780 nm on potassium titanyl phosphate crystals pumped by a 390-nm pulsed laser. The source provides a coherence time of tens of picoseconds, and it is shown to have the potential to be applied in long-distance quantum interference. Furthermore, we experimentally demonstrate the Hong-Ou-Mandel interference between two photon sources with visibility exceeding the classical limit.

Hybrid quantum systems based on magnetic platforms have witnessed the birth and fast development of quantum magnonics. Until now, most studies have relied on magnetic excitations in low-damping magnetic insulators, particularly yttrium iron garnet, while a large class of magnetic systems is ruled out in this interdisciplinary field. Here we propose the generation of a magnon bundle in a hybrid magnet-qubit system, where two or more magnons with the same energy and angular momentum are emitted simultaneously. By tuning the driving frequency of the qubit to match the detuning between magnon and qubit mode, one can effectively generate a magnon bundle via super-Rabi oscillations. In contrast with general wisdom, magnetic dissipation plays an enabling role in generating the magnon bundle, where the relaxation time of magnons determines the typical time delay between two successive magnons. The maximal damping that allows an antibunched magnon bundle can reach the order of 0.1, which may break the monopoly of low-dissipation magnetic insulators in quantum magnonics and enables a large class of magnetic materials for their applications in quantum manipulation. Further, our findings may provide a scalable and generic platform to study multimagnon physics and benefit the design of magnonic networks for quantum information processing.

The implementation of a universal quantum processor still poses fundamental issues related to error mitigation and correction, which demand investigation of also platforms and computing schemes alternative to the main stream. A possibility is offered by employing multilevel logical units (qudits), naturally provided by molecular spins. Here we present the blueprint of a molecular spin quantum processor consisting of single molecular nanomagnets, acting as qudits, placed within superconducting resonators adapted to the size and interactions of these molecules to achieve a strong single spin-to-photon coupling. We show how to implement a universal set of gates in such a platform and to readout the final qudit state. Single-qudit unitaries (potentially embedding multiple qubits) are implemented by fast classical drives, while an alternative scheme is introduced to obtain two-qubit gates via resonant photon exchange. The latter is compared to the dispersive approach, finding in general a significant improvement. The performance of the platform is assessed by realistic numerical simulations of gate sequences, such as Deutsch-Josza and quantum simulation algorithms. The very good results demonstrate the feasibility of the molecular route towards a universal quantum processor.

Neutral shallow donors ($D^0$) in $\mathrm{Zn}\mathrm{O}$ have emerged as a promising candidate for solid-state spin qubits. Here we report on the formation of $D^0$ in $\mathrm{Zn}\mathrm{O}$ via implantation of $\mathrm{In}$ and subsequent annealing. The implanted $\mathrm{In}$ donors exhibit optical and spin properties on par with those of in situ–doped donors. The inhomogeneous linewidth of the donor-bound-exciton transition is less than 10 GHz, comparable to the optical linewidth of in situ $\mathrm{In}$. Longitudinal spin relaxation times ($T_1$) exceed reported values for in situ $\mathrm{Ga}$ donors, indicating that residual $\mathrm{In}$-implantation damage does not degrade $T_1$. Two-laser Raman spectroscopy of the donor spin reveals the hyperfine interaction of the donor electron with the spin-9/2 $\mathrm{In}$ nuclei. This work is an important step toward the deterministic formation of $\mathrm{In}$-donor qubits in $\mathrm{Zn}\mathrm{O}$ with optical access to a long-lived nuclear-spin memory.

Gate-defined quantum dots (QDs) have appealing attributes as a quantum computing platform. However, near-term devices possess a range of possible imperfections that need to be accounted for during the tuning and operation of QD devices. One such problem is the capacitive crosstalk between the metallic gates that define and control QD qubits. A way to compensate for the capacitive crosstalk and enable targeted control of specific QDs independent of coupling is by the use of virtual gates. Here, we demonstrate a reliable automated capacitive coupling identification method that combines machine learning with traditional fitting to take advantage of the desirable properties of each. We also show how the cross-capacitance measurement may be used for the identification of spurious QDs sometimes formed during tuning experimental devices. Our systems can autonomously flag devices with spurious dots near the operating regime, which is crucial information for reliable tuning to a regime suitable for qubit operations.

Lithium niobate (${\mathrm{Li}\mathrm{Nb}\mathrm{O}}_3$, LN) devices play critical roles in quantum information processing. However, for special applications like quantum key distribution (QKD), the characteristics of materials and devices and their impact on practical systems must be intensively explored. For the first time, we reveal that the photorefractive effect in LN can be utilized as a potential loophole to carry out malicious attacks by the eavesdroppers. We take a commercial LN-based variable optical attenuator as an example to demonstrate the method we name the induced-photorefraction attack (IPA) and propose two techniques to enable controllable attacks. Our results show that eavesdroppers can fulfill an efficient source-side attack by injecting an optimized irradiation beam with only several nanowatts, which is realistic when accessing commercial fiber channels. These measures and techniques can be employed for all individual and on-chip LN devices. Our work opens a new window for the security research of real-life QKD and contributes to designing QKD systems with higher security.