Voltage-controlled skyrmion-based nanodevices for neuromorphic computing using a synthetic antiferromagnet

Area of Science:

• Spintronics
• Neuromorphic Computing
• Materials Science

Background:

• Spintronics offers potential for high-speed, dense, and low-dissipation neuromorphic computing systems.
• Existing neuromorphic devices face challenges in energy efficiency and functional mimicry.

Purpose of the Study:

• To propose and demonstrate an ultralow-dissipation skyrmion-based nanodevice for neuromorphic computing applications.
• To utilize synthetic antiferromagnets and piezoelectric substrates for efficient synaptic and neuronal functions.

Main Methods:

• Fabrication of a nanodevice using a synthetic antiferromagnet (SAF) heterostructure on a piezoelectric substrate.
• Generation and manipulation of skyrmions/skyrmion bubbles in the SAF layer via electric field application.
• Measurement of resistance variations in a magnetic tunneling junction to emulate synaptic plasticity and neuronal firing.

Main Results:

• Demonstrated continuous transition between large and small skyrmions by manipulating interlayer antiferromagnetic coupling with a weak electric field.
• Achieved emulation of synaptic potentiation/depression and leaky-integral-and-fire neuron function.
• Attained ultralow energy consumption of 0.3 fJ per operation.

Conclusions:

• The proposed skyrmion-based nanodevice offers a viable pathway towards ultralow power neuromorphic computing.
• The device integrates spintronic and piezoelectric functionalities for efficient brain-inspired computing.
• This research advances the development of energy-efficient artificial intelligence hardware.