Artificial neurons based on antiferromagnetic auto-oscillators as a platform for neuromorphic computing

Spiking artificial neurons emulate the voltage spikes of biological and constitute building blocks a new class energy efficient, neuromorphic computing systems. Antiferromagnetic materials can, in theory, be used to construct spiking neurons. When configured as neuron, magnetization antiferromagnetic has an effective inertia that gives them intrinsic characteristics closely resemble neurons, contrast with conventional It is shown here have spike duration on order picoseconds, power consumption about 10 −3 pJ per synaptic operation, built-in features directly including response latency, refraction, inhibition. also demonstrated interconnected into physical neural networks can perform unidirectional data processing even for passive symmetrical interconnects. The flexibility illustrated by simulations simple circuits realizing Boolean logic gates controllable memory loops.