Category Physics

New memristor training method slashes AI energy use by six orders of magnitude

New memristor training method slashes AI energy use by six orders of magnitude
The implementation of SRNet ×2 on memristor hardware. Credit: Nature Communications (2025). DOI: 10.1038/s41467-025-66240-7

In a Nature Communications study, researchers from China have developed an error-aware probabilistic update (EaPU) method that aligns memristor hardware’s noisy updates with neural network training, slashing energy use by nearly six orders of magnitude versus GPUs while boosting accuracy on vision tasks. The study validates EaPU on 180 nm memristor arrays and large-scale simulations.

Analog in-memory computing with memristors promises to overcome digital chips’ energy bottlenecks by performing matrix operations via physical laws. Memristors are devices that combine memory and processing like brain synapses.

Inference on these systems works well, as shown ...

Read More

Soft, 3D transistors could host living cells for bioelectronics

Soft, 3D transistors may upend semiconductor device design, transform bioelectronics
From 2D rigid electronics to 3D soft electronics: Increasing the dimensionality of transistors with 3D hydrogel semiconductors. Credit: Shiming Zhang

New research from the WISE group (Wearable, Intelligent, Soft Electronics) at The University of Hong Kong (HKU-WISE) has addressed a long-standing bioelectronic challenge: the development of soft, 3D transistors.

This work introduces a new approach to semiconductor device design with transformative potential for bioelectronics. It is published in Science.

Led by Professor Shiming Zhang from the Department of Electrical and Electronic Engineering, Faculty of Engineering, the research team included senior researchers who joined HKU-WISE from the University of Cambridge and the University of Chicago, together with HKU Ph.D...

Read More

New state of matter discovered in a quantum material

three people in a physics lab, in the center a technical device with metallic spiral
In the Microkelvin Lab
Silke Bühler-Paschen (left), Diego Zocco and Diana Kirschbaum

At TU Wien, researchers have discovered a state in a quantum material that had previously been considered impossible. The definition of topological states should be generalized.

The work is published in Nature Physics.

Quantum physics tells us that particles behave like waves and, therefore, their position in space is unknown. Yet in many situations, it still works remarkably well to think of particles in a classical way—as tiny objects that move from place to place with a certain velocity.

When physicists describe how electric current flows through metals, for example, they imagine electrons racing through the material and being accelerated or deflected by electromagnetic fields.

Even mo...

Read More

Telecommunications beyond 6G: the first standalone spinwave chip with a built-in magnetic field

Cross-section and top view of the magnonic device with integrated micromagnets
Cross-section and top view of the magnonic device with integrated micromagnets

Credit
Politecnico di Milano

The Politecnico di Milano has created the first integrated and fully tunable device based on spin waves, opening up new possibilities for the telecommunications of the future, far beyond current 5G and 6G standards. The study, published in the journal Advanced Materials, was conducted by a research group led by Riccardo Bertacco of the Department of Physics of the Politecnico di Milano, in collaboration with Philipp Pirro of Rheinland-Pfälzische Technische Universität and Silvia Tacchi of Istituto Officina dei Materiali—CNR-IOM.

Magnonics is an emerging technology that uses spin waves—collective excitations of electronic spins in magnetic materials—as an alternative to elect...

Read More