Biologically Realistic Dynamics for Nonlinear Classification in CMOS+X Neurons
2026-04-03 • Neural and Evolutionary Computing
Neural and Evolutionary Computing
AI summaryⓘ
The authors investigate how spiking neural networks can do complex calculations using compact hardware that combines standard electronics with magnetic components. They study a special neuron made from a magnetic tunnel junction and a transistor, showing through simulations that it can solve problems like XOR by using three key features: a threshold to decide if the neuron fires, timing delays that affect when it fires, and a pause period after firing. This suggests that the magnetic behavior in these devices can help build efficient brain-like chips without extra complicated circuits.
spiking neural networksmagnetic tunnel junctionCMOSNMOS transistornonlinear computationthreshold activationresponse latencyabsolute refractorinessXOR problemneuromorphic hardware
Authors
Steven Louis, Hannah Bradley, Artem Litvinenko, Cody Trevillian, Darrin Hanna, Vasyl Tyberkevych
Abstract
Spiking neural networks encode information in spike timing and offer a pathway toward energy efficient artificial intelligence. However, a key challenge in spiking neural networks is realizing nonlinear and expressive computation in compact, energy-efficient hardware without relying on additional circuit complexity. In this work, we examine nonlinear computation in a CMOS+X spiking neuron implemented with a magnetic tunnel junction connected in series with an NMOS transistor. Circuit simulations of a multilayer network solving the XOR classification problem show that three intrinsic neuronal properties enable nonlinear behavior: threshold activation, response latency, and absolute refraction. Threshold activation determines which neurons participate in computation, response latency shifts spike timing, and absolute refraction suppresses subsequent spikes. These results show that magnetization dynamics of MTJ devices can support nonlinear computation in compact neuromorphic hardware.