Neuromorphic spintronics is a booming new field worldwide which utilize the spin degrees of freedoms to fabricate artificial nanodevices which mimic the functionality of our brain. The manipulation of spin textures and dynamics can not only serve as biological synapses and neurons, but also develop new computing algorism and schemes in very efficient ways.
In our group, we focus on several aspects: (1) In-memory computing using magnetic oxide nanostructures; (2) Reservoir computing using artificial spin ice system; (3) Neural networks with ultrafast natural learning capabilities using magnetic textures; (4) High efficiency probabilistic computing using correlated magnetic oxides; (5) Artificial human retina using magnetic neural networks.
References:
1.Q. Shi et al., “An electric-field-controlled high-speed coexisting multibit memory and Boolean logic operations in manganite nanowire via local gating” Adv. Electron. Mater. 5, 1900020 (2019)
2.C. Niu et al., “Implementation of artificial neurons with tunable width via magnetic anisotropy” Appl. Phys. Lett. 119, 204101 (2021).
3.W. J. Hu et al., "Distinguishing artificial spin ice states using magnetoresistance effect for neuromorphic computing" Nature Communications 14,2562 (2023)
4.时倩等，一种集多值存储和逻辑运算于一体的器件单元及操作方法，专利号：ZL 2018 1 0111023.8
5.“低维体系自旋态的调控及其在未来人工智能器件中的应用”，国家重点研发计划，在研，2200万。
