Advanced quantum technologies最新文献

The relaxation rate is a fundamental parameter that characterizes the dynamics of atomic ensembles and plays a critical role in the performance of atomic magnetometers. In this study, a transient dynamics model for the NMOR magnetometer is developed based on multipole moment theory, revealing that the decay rate of the transient response to a step magnetic field from zero to certain value is determined by both transverse and longitudinal relaxation rates, whereas that to a step magnetic field from certain value to zero depends solely on the longitudinal relaxation rate. This distinction enables the independent determination of the two relaxation rates. To validate the proposed method, transverse and longitudinal relaxation rates are experimentally measured under various laser powers, which are consistent with the Lorentzian linewidths extracted from magnetic resonance curve fitting. This study provides valuable insights into atomic transient dynamics and contributes to the fast measurement of relaxation rates in NMOR magnetometers.

A silicon crystal hosts an ion implanted phosphorus atom with a nuclear spin surrounded by its simulated donor electron wavefunction in a quantum superposition between the donor atom and a potential well from an isolated surface electrode. A Moiré pattern arises from periodic localisation of the wavefunction around silicon atoms in the crystal that are spaced 0.5 nm apart. More in article number e00675, David N. Jamieson and co-workers.

The controlled integration of quantum dots (QDs) as single-photon emitters into quantum light sources is essential for the implementation of large-scale quantum networks. In this study, we employ the deterministic in situ electron-beam lithography (iEBL) nanotechnology platform to integrate individual QDs with high accuracy and process yield into a circular Bragg grating (CBG) resonators. Notably, CBG devices comprising just 3 to 4 rings exhibit photon extraction efficiencies comparable to those of structures with more rings. This facilitates faster fabrication, reduces the device footprint, and enables compatibility with electrical contacting. To demonstrate the scalability of this process, we present results of 95 optically active QD-CBG devices fabricated across two lithography sessions. These devices exhibit bright, narrow-linewidth single-photon emission with excellent optical quality. To evaluate QD placement accuracy, we apply a powerful characterization technique that combines cathodoluminescence (CL) mapping and scanning electron microscopy. Statistical analysis of these devices reveals that our iEBL approach enables high alignment accuracy and a process yield of over $��>�90�\%�$ across various CBG geometries. Our findings highlight a reliable route toward the scalable fabrication of high-performance QD-based single-photon sources for use in photonic quantum technology applications.

The Hyperband-QNN algorithm achieves adaptive customization of quantum neural network structure by skilfully integrating the essence of neural networks and the Hyperband optimization algorithm, perfectly matching the needs of various specific tasks. This groundbreaking method not only opens up a new path for the design of quantum neural networks but also sets a model for the deep integration of quantum computing and traditional computing. More in article number 2400178, Zheng Shan and co-workers.

Quantum-noise-driven diffusion models are proposed here as a novel class of quantum generative AI algorithms. These models aim to exploit the intrinsic noise of currently available quantum processing units, not as an issue to be solved by quantum error mitigation and correction, but instead as a beneficial resource to generate artificial, classical or quantum, data sampled from some unknown and usually very complex probability distribution that could be difficult or even impossible to sample from via classical computers. More in article number 2300401, Filippo Caruso and co-workers.

In article number 2400603, Alexey Melnikov and co-workers present a method to enhance the generalization capability of quantum machine learning models through controllable noise regularization. The cover visualizes a quantum circuit in which quantum gates are interleaved with engineered noise channels of tunable strength λ. This structured injection of noise acts as an implicit regularizer, stabilizing the training process and improving the model's predictive performance on previously unseen data.

This image illustrates the process of Rydberg-based quantum-enhanced simulated annealing (QESA), which achieves faster computational performance than classical simulated annealing (SA). Harnessing Rydberg interactions tailored to the maximum independent set problem, QESA demonstrates a clear quantum advantage in heuristic optimization, pointing to a promising route for tackling complex combinatorial problems more efficiently than classical methods. More in article number 2500070, Jaewook Ahn and co-workers.