Voltage-controlled skyrmion-based nanodevices forneuromorphic computing in a synthetic antiferromagnet

Spintronics exhibits significant potential in neuromorphic computing system with high speed, highintegration density, and low dissipation. In this letter, we propose an ultralow-dissipationskyrmion-based nanodevices for neuromorphic computing composed of a syntheticantiferromagnet (SAF) and a piezoelectric substrate. Skyrmions/skyrmion bubbles can begenerated in the upper layer of SAF with a weak anisotropy energy (Ea). Under a weak electricfield on the heterostructure, the interlayer antiferromagnetic coupling can be manipulated, givingrise to a continuous transition between a large skyrmion bubble and a small skyrmion. This thusinduces the variation of the resistance of a magnetic tunneling junction that can mimic thepotentiation/depression of a synapse and the leaky-integral-and-fire function of a neuron at a costof a very low energy consumption of 0.3 fJ. These results pave a way to ultralow powerneuromorphic computing applications.