A new route to optimize AI hardware: Homodyne gradient extraction

November 24, 2025 by K.w. Wesselink – Schram, University of Twente

Collected at: https://techxplore.com/news/2025-11-route-optimize-ai-hardware-homodyne.html

A team led by the BRAINS Center for Brain-Inspired Computing at the University of Twente has demonstrated a new way to make electronic materials adapt in a manner comparable to machine learning. Their study, published in Nature Communications, introduces a method for physical learning that does not require software algorithms such as backpropagation. Backpropagation—the optimization method popularized in the 1980s by Nobel Prize winner Geoffrey Hinton and colleagues—is at the heart of today’s AI revolution.

Human brain: Energy consumption of a light bulb

Modern AI relies on backpropagation running on powerful digital computers. While this approach delivers remarkable performance, it is also extremely energy hungry. The human brain, by contrast, achieves similar tasks with just the energy of a light bulb. Neuromorphic hardware offers a path toward far greater efficiency, but cannot easily be trained using backpropagation.

The Twente team’s approach, called homodyne gradient extraction (HGE), makes it possible to find the optimum operating point of physical neural networks directly in hardware, without any software-based optimization. While external perturbations are still applied, the optimization itself takes place in the device, eliminating the need for digital computers and backpropagation algorithms.

“This opens the door to stand-alone optimization of physical neural networks, offering a path towards energy-efficient, adaptive hardware,” says Prof. Wilfred van der Wiel, co-director of BRAINS. Potential applications include smart sensors that adapt on the spot and brain-inspired computers designed for sustainable, low-energy information processing.

More information: Marcus N. Boon et al, Gradient descent in materia through homodyne gradient extraction, Nature Communications (2025). DOI: 10.1038/s41467-025-65155-7

Journal information: Nature Communications