Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
This Perspective explores the potential of directly mapping computational problems in machine learning to materials and device properties, and proposes metrics to facilitate comparisons between different solutions to machine learning tasks.
This Review examines switching mechanisms in memristive devices based on van der Waals materials, and explores the advantages such devices offer and the challenges that must be faced for them to be of use in next-generation electronic and computing applications.
The use of topological spin structures is restricted by their limited scale, thermal stability or magnetic field requirements. A high-magnetic-field-assisted growth approach overcomes these limitations, enabling the construction of millimetre-scale meron lattices. These lattices were used to demonstrate chirality transfer from topologically protected quasiparticles to electrons and then photons.
Physically unclonable functions that are based on magnetic random-access memory, and integrated with complementary metal–oxide–semiconductor circuitry, can be used to create secure and efficient compute-in-memory macros for edge computing.
A technique based on a scanning tunnelling microscope can provide simultaneous control, visualization and spectroscopic characterization of quantum states with atomic resolution.
A magnetic random-access memory device that has an antiferromagnetic material as its storage element can be electrically read using ferromagnetic tunnelling.
Stacking a bilayer of chromium triiodide, a layered antiferromagnet, onto another with a twist angle gives rise to a moiré magnet with rich magnetic phases, including ferromagnetic and antiferromagnetic orders. The magnetic orders can be controlled through the twist angle, temperature and electrical gating, with the system also showing voltage-assisted magnetic switching.
The negative differential capacitance (NDC) of ferroelectrics could be used to reduce the energy consumption of ultra-scaled logic devices. An NDC phenomenon in ultrathin ferroelectric zirconium-doped hafnia is demonstrated. Field-effect transistors incorporating this ferroelectric in the gate stack display enhanced on-currents and reduced off-currents compared with conventional analogues, as well as tunable and enduring NDC.
Despite advances in speech processing systems, such as those used in voice-controlled devices, human hearing still outperforms technical systems in noisy and variable environments. To close this gap, a bioinspired acoustic sensor with integrated signal processing was developed — the dynamic microelectromechanical system (MEMS)-based cochlea.
A spoof surface plasmonic neural network with programmable weights and activation functions was proposed, which has the potential to achieve processing speeds close to the speed of light. This neural network was used to create a wireless communications system that can detect and process electromagnetic waves.
An elastic conductive ink — which is made of conductive fillers suspended in an emulsified elastomer matrix — can be used to print three-dimensional elastic conductors.
By selectively engineering the surface roughness of micro-light-emitting-diode chips, and thus the strength of the van der Waals forces that bond them to a substrate, large-area displays can be created via a fluidic-assisted transfer method.
This Review examines the development of cryogenic memory technologies—including non-superconducting memories, superconducting memories and hybrid memories—and their potential application in superconducting single-flux quantum circuits and quantum computers.
Multilayer hexagonal boron nitride can be synthesized over large areas and used to enhance mobility in graphene heterostructures, illustrating the potential of the material as an insulator in commercial two-dimensional electronics.
Conventional dielectric layers used in stretchable electronics are solution-processed, thick and have poor electrical performance compared with rigid, inorganic dielectrics. A stretchable nanometre-thick gate dielectric layer has been produced using large-area vacuum deposition. This material has excellent electrical, mechanical and chemical properties and could facilitate the development of high-performance wearable devices.