A memory technology that combines the functions of memristors and ferroelectric capacitors in a single stack can be used for on-chip training and inference of artificial neural networks.

The electrical and thermal transport properties of amorphous indium tin oxide can be enhanced by using a silicon carbide substrate, leading to indium tin oxide power amplifiers with an output power density of 0.69 W mm⁻¹ and a power-added efficiency of 24.1% at 12 GHz.

An embedded microfluidic cooling approach—consisting of an integrated manifold layer, jet plate and sawtooth-shaped microchannels—can enable high heat flux dissipation up to 3,000 W cm⁻² with a pumping power under 1 W cm⁻².

Biocompatible and stretchy organic field-effect transistors can be created using a vulcanized blend film of semiconducting nanofibres and a medical-grade elastomer.

Sixteen-element integrated microwave transmitters that are fabricated on four-inch monolayer molybdenum disulfide wafers offer a low power consumption of 3.2 μW.

A soft, finger-worn haptic interface consisting of four serpentine-structured shape memory alloy actuators can implement 11 distinct patterns of tactile sensations composed of shear, normal and circular rubbing forces.

Photodetectors based on emerging semiconductors such as quantum dots and perovskites have been under development for over two decades, but the use of diverse characterization methods and set-ups make it increasingly difficult to compare device performance. A more standardized approach to external quantum efficiency characterization is now essential.

Neuromorphic hardware has typically focused on accelerating vector–matrix multiplication, but broader and more disruptive approaches will be required to reimagine AI hardware.

The development of a universal quantum electrical standard would help deliver a more accessible measurement system.