Nanoscale neural network using non-linear spin-wave interference

Wave based computing has sparked much interest for neuromorphic computing due to the inherent interconnectedness of such wave based approaches. Here, Papp, Porod and Csaba show how neural networks can be implemented using spin-waves, taking advantage of spin-waves intrinsic non-linearity. We demonstrate the design of a neural network hardware, where all neuromorphic computing functions, including signal routing and nonlinear activation are performed by spin-wave propagation and interference. Weights and interconnections of the network are realized by a magnetic-field pattern that is applied on the spin-wave propagating substrate and scatters the spin waves. The interference of the scattered waves creates a mapping between the wave sources and detectors. Training the neural network is equivalent to finding the field pattern that realizes the desired input-output mapping. A custom-built micromagnetic solver, based on the Pytorch machine learning framework, is used to inverse-design the scatterer. We show that the behavior of spin waves transitions from linear to nonlinear interference at high intensities and that its computational power greatly increases in the nonlinear regime. We envision small-scale, compact and low-power neural networks that perform their entire function in the spin-wave domain.

Automated summary

Wave based computing in neuromorphic systems utilizes spin-waves for non-linear operations, with signal routing and weight interconnections achieved through interference of scattered waves. Training the network involves finding the desired input-output mapping field pattern, and the computational power greatly increases in the non-linear interference regime.