Skyrmion-Magnetic Tunnel Junction Synapse With Long-Term and Short-Term Plasticity for Neuromorphic Computing

Magnetic skyrmion-based data storage and unconventional computing devices have gained increasing attention due to their topological protection, small size, low driving current. However, skyrmion creation, deletion, motion are still being studied. In this study, we propose a neuromorphic magnetic tunnel junction (MTJ) device with both long- short-term plasticity (LTP STP) (mixed synaptic plasticity). We showed that could be controlled by field, spin-orbit torque (SOT), the voltage-controlled anisotropy (VCMA) switching mechanism. LTP depends on density is manipulated SOT field while STP VCMA. The property of was utilized for static image recognition. By incorporating feature, additional temporal filtering ability adapt dynamic environment. skyrmions were conserved confined nanotrack minimize nucleation energy. synapse trained tested emulating deep neural network. observed when increased, inference accuracy improved: 90% achieved system at highest density. further demonstrated environment learning capabilities proposed device.