UC San Diego researchers say a quantum material can be pushed into states that make a tiny electronic device behave more like a learning neuron.

In a study highlighted October 13, the team watched a vanadium-dioxide neuristor while electricity flowed through it using coherent X-ray diffraction. A neuristor is an electronic device designed to imitate some behavior of biological neurons.

UC San Diego’s Jacobs School has previously highlighted the wider push to scale neuromorphic computing for more energy-efficient AI hardware.

The researchers found that repeated above-threshold electrical pulses reorganized nanoscale domains into a jammed, glass-like state. Once jammed, the device switched more predictably. Lower-current pulses could erase that short-term structural memory and restore more random switching. The peer-reviewed Nature Communications paper describes this as a possible route toward learning directly inside hardware.

The Royal Institution explains why neuromorphic computing is being explored as a lower-power alternative to conventional computing architectures.

This is not a material that “thinks” like a person. The learning and forgetting refer to controllable physical switching behavior inside the device. But that distinction is exactly why the result matters: future AI hardware may gain useful memory from the physics of the material itself instead of constantly moving data between separate processors and memory. That direction complements BitcoinVersus.Tech’s coverage of new physical-AI processors, lower-power dataflow chips, and the rising power demand behind cheaper AI.

BitcoinVersus.Tech Editor’s Note:
We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. If you would like to support to help further secure the integrity of our research initiatives, please donate using the QR code provided

For best QR-code scanning results, please scan this donation code from a desktop or larger screen. Mobile screens may prevent camera recognition.