Search

Landauer’s principle connects entropy and energy dissipation in non-equilibrium processes. An experiment now uses this principle to measure entropy production in a Bose gas to resolve contributions from correlations and dissipation.

The authors report the implementation of a Transformer-based model on the same architecture used in Large Language Models in a 14nm analog AI accelerator with 35 million Phase Change Memory devices, which achieves near iso-accuracy despite hardware imperfections and noise.

Large volumes of true random numbers are needed for increasing requirements of secure data encryption. Here the authors use the stochastic nature of DNA synthesis to obtain millions of gigabytes of unbiased randomness.

Froths and foams are complex structures, particularly those that disappear irreversibly. Superconducting froth, however, can be reversibly controlled by several external parameters, so it may help quantify froth dynamics across different systems.

Diffusion models have shown promise in content generation, however, like many forms of generative AI, this comes with high computation cost, exacerbated by standard von Neuman computing architectures. Here, Cheng et al present a magnetoelectric memory for in-memory computing and demonstrate diffusion-based image generation on an 80×80 array.

Human transplantation with allogeneic donor organs results in non-matching of MHC and differential presentation of T cell antigens. Here the authors show that in a lung transplanted SARS-CoV-2 infected patient T cell responses generated from the host may not be able to recognise infected cells within the graft and this may contribute to virus persistence.

Genome-wide association meta-analyses of attention-deficit/hyperactivity disorder symptom measures and clinical diagnoses identify new risk loci, highlight putative effector genes and yield improved polygenic risk scores.

Conventional mechanoreceptors are limited by the trade-off between sensitivity and stability. Here, the authors present an artificial tactile neuromorphic device with a mechanical sensing unit based on a non-faradaic junction and a neuromorphic computing unit, enabling accurate object recognition.

This study demonstrates stable photoelectricity-induced halide segregation in epitaxial mixed-halide perovskite microwire networks on mica, verified by in-situ observations. Dynamic segregation-recovery processes and reconfigurable self-powered photoresponse reveal physical computing potential.

The authors introduce a novel method (AC-QUDIT) for achieving high-fidelity transport of quantum wavepackets in the presence of dissipation and leakage to the continuum. This method provides an alternative to shortcuts to adiabaticity (or counter-diabatic driving), with more general applicability and better performance.

The security of DIQKD is difficult to prove, as one needs to take into account every possible attack strategy. Here, the authors develop a method to determine the entropy of a system as the sum of the entropies of its parts. Applied to DIQKD, this implies that it suffices to consider i.i.d. attacks.

This study shows a viable pathway to the efficient deployment of state-of-the-art large language models using mixture of experts on 3D analog in-memory computing hardware.

The Gaussian minimum entropy conjecture—a long-standing open question—has now been proved for single-mode phase-insensitive bosonic Gaussian channels. This establishes the ultimate achievable bit rate under an energy constraint and provides long-awaited proof that the single-letter classical capacity of these channels is additive.

Proseg is a segmentation approach for single-cell spatially resolved transcriptomics data that uses unsupervised probabilistic modeling of the spatial distribution of transcripts to accurately segment cells without the need for multimodal staining.

The presence of correlations can strongly affect the evolution of a quantum system. Here, the authors directly observe differences in the dynamics of two spins-1/2 systems in an NMR setup depending on the correlations of the initial state, including differences in energy flow and mutual information.

By using a resonant sensor to couple two radio-frequency parametric oscillators behaving as Ising spins, a passive wireless device can implement programmable temperature threshold sensing.

Advancing single-atom quantum information processing necessitates a deep understanding of electron and nuclear spin dynamics. Here, using pump-probe spectroscopy, the authors detect the coherent dynamics of a nuclear and electron spin of a single hydrogenated Ti atom on MgO surface.

The renormalization group is a powerful tool to study the universal properties of physical systems. A diffusion-based renormalization scheme now enables the study of scale invariance and universality in higher-order complex networks.

A hybrid system that combines a von Neumann machine with a computational memory unit can offer both the high precision of digital computing and the energy/areal efficiency of in-memory computing, which is illustrated by accurately solving a system of 5,000 equations using 998,752 phase-change memory devices.

Chirality-induced quantum non-reciprocity of cross-channel correlations is demonstrated in a rubidium vapour system by flipping the flow direction of one of the circularly polarized laser beams. It can be extended to multicolour sidebands with Floquet engineering.

The authors developed a neuromorphic chip with on-chip learning and support for diverse memory devices. It bridges brain-inspired computing and emerging tech, enabling efficient, flexible testing and advancing next-gen neuromorphic architectures.

Using well-calibrated statistical methods the authors jointly analyze Undiagnosed Diseases Network genomes, identifying known and novel disease genes. Software is publicly available to support future cross-cohort rare disease discovery efforts.

Recent studies have reported miniaturized spectrometers based on van der Waals heterostructures. Here, the authors demonstrate multifunctional SnS₂/ReSe₂ heterojunction spectrometers providing photodetection, spectrum reconstruction, spectral imaging, long-term image memory, and signal processing capabilities.

Irreversible computation cannot be performed without a work cost, and energy dissipation imposes limitations on devices' performances. Here the authors show that the minimal work requirement of logical operations is given by the amount of discarded information, measured by entropy.

Continuous variable quantum key distribution allows secure communication that is more robust against channel losses than discrete approaches, yet is strongly affected by noise. Madsenet al.devise a continuous scheme for modulated entangled states that is more tolerant to noise and loss than other protocols.

The Heisenberg uncertainty principle bounds the uncertainties about the outcomes of two incompatible measurements on a quantum particle. This bound, however, changes if a memory device is involved that stores quantum information. New work now extends the uncertainty principle to include the case of quantum memories, and should provide a guide for quantum information applications.

The scaling of entanglement entropy and mutual information is key for the understanding of correlated states of matter. An experiment now reports the measurement of von Neumann entropy and mutual information in a quantum field simulator.

Continuous-variable remote state preparation in the microwave domain would allow to leverage the superconducting technology for quantum networks applications. Here, the authors show how to deterministically prepare squeezed Gaussian states across 35 cm using previously shared entanglement.

One particularly useful feature of van der Waals materials is the ability to combine layers of different materials into a single heterostructure, which can have superior properties than any of the constituent materials alone. Here, Cheng et al. combine two interlayer-antiferromagnetic chromium trihalides, CrI₃ and CrCl₃ in close proximity, and demonstrate ferromagnetic coupling between them.

Computing approaches in the optical domain would allow for ultra-fast signaling and ultra-high bandwidth capabilities. Here, Feldmann et al. demonstrate a photonic abacus, which provides multistate compute-and store operation by integrating phase-change materials with nanophotonic chips.

Device-independent quantum key distribution aims at the ultimate quantum-based unconditional security, but current protocols’ rates are quite far from anything practical. The authors’ protocol narrows this gap by using two randomly chosen key generating bases instead of one.

The exact crystal chemistry of perovskites containing 5f elements has remained a matter of debate. Here, the authors synthesise a BaO-deficient Pu-based perovskite with a composition close to Ba3PuO6 and study the Pu oxidation state, the crystal structure and Ba/Pu ratio, reporting on thermodynamic and magnetic properties and offering new insights into Ba-Pu-O solid-state chemistry.