The relentless march of artificial intelligence (AI) and machine learning is fueling a demand for faster, more energy-efficient hardware. Inspired by the intricate workings of biological neural networks, engineers are increasingly focused on creating hardware that mirrors the human brain’s architecture and functionality.

Neuromorphic hardware,as this brain-inspired technology is known,typically incorporates components designed to replicate the functions of biological neurons and synapses.

New Artificial Neuron Shows Promise

A team of researchers led by Yin Wang and Gou have engineered an artificial neuron that combines dynamic random-access memory (DRAM) with molybdenum disulfide (MoS2) circuits. This innovative design allows for both “light coding and visual adaptation,” according to the authors. “We also use the neuron module to simulate a bioinspired neural network model for image recognition.”

We also use the neuron module to simulate a bioinspired neural network model for image recognition.

The artificial neuron developed by Wang, Gou and their colleagues has demonstrated significant potential, especially in the energy-efficient implementation of computer vision and image recognition models. The researchers believe that their newly developed device could serve as a foundation for future bio-inspired computing systems, with potential applications across a range of computational tasks.