Antiferromagnetic Skyrmion Based Energy-Efficient Integrate-Fire Neuron Device

Antiferromagnetic (AFM) skyrmion technology holds great promise for the development of the next-generation spintronics-based neuromorphic computing due to its unique features. They offer negligible stray fields, resilience to external magnetic fields, ultrafast dynamics, and no skyrmion Hall effect (SkHE) due to zero net topological charge. These advantages make AFM skyrmions a more viable alternative for future spintronic applications compared to ferromagnetic (FM) skyrmions. In this study, a neuron device based on AFM skyrmion is proposed, which demonstrates the integrate-fire (IF) functionality by utilizing a previously proposed AFM skyrmion-based diode device and the interskyrmion repulsion effect on the nanotrack. The threshold of the device, which is determined by the maximum number of skyrmions pinned near the device window, can be adjusted by manipulating the applied current density. The device generates an output signal when the primary skyrmion overcomes the device window and reaches the detection region. By increasing the threshold, the device’s operating current density can be reduced, leading to improved energy efficiency. It is observed that the threshold current density is 4.6 GAm
$^{-{2}}$
, 2.3 GAm
$^{-{2}}$
, and 1.5 GAm
$^{-{2}}$
for one, two, and three skyrmions, respectively, on a nanotrack. The minimum overall energy dissipation is estimated at 6.285 fJ. The proposed neuron device offers a 68% reduction in the energy dissipation when the number of skyrmions on a nanotrack is increased from 1 to 3, thus indicating the potential for developing energy-efficient devices in AFM spintronics for neuromorphic computing.