New hardware chips use the physical properties of glass to perform division and addition at the speed of thought.
Traditional AI hardware relies on digital transistors to simulate the math required for neural networks. These Ovonic threshold switches physically mimic the shunting inhibition found in biological neurons. This allows for analog computing that is orders of magnitude more energy efficient than current GPUs. By using the natural behavior of materials, we can build AI that runs on a fraction of the power of a standard lightbulb. This technology could finally bring high-performance AI to tiny, battery-powered edge devices. We are moving from simulating brains to building them out of physical matter.
Neuronal arithmetic operators based on Ovonic threshold switches (OTS) for biologically inspired analog computing
arXiv · 2604.27650
Biological neurons perform arithmetic computations - including additive integration and divisive gain modulation - through synaptic conductance changes and shunting inhibition, enabling context-dependent information processing that far exceeds simple threshold-and-fire models. Replicating these capabilities in compact hardware remains a fundamental challenge for neuromorphic engineering. Here, we demonstrate artificial neuron circuits based on Ovonic threshold switches (OTS) that physically impl