Perspective on spin-based wave-parallel computing

Abstract

Waves exhibit unique characteristics, such as diffraction and interference, which distinguishes them from the particle nature of electrons currently used for binary and sequential data processing and storage. In the solid state, wave properties can be found in electron spin waves in semiconductors or magnons in magnetic materials. These are useful for communication, processing and storage, and allow multiplexing of the information. Given this perspective, after introducing the information theory of wave-parallel computing and arguing the fundamental properties necessary for implementation with wave-based information carriers, we specifically examine how electron spin waves and magnons can be used as information carriers for processing and storage. Then, after explaining the fundamental physics of the electron spin wave based on the persistent spin helix state, we assess the potential of magnon-assisted magnetization switching for realizing the selective writing and reading of multiplexed information. Ferromagnet/semiconductor hybrid structures are emphasized as a platform for generating and controlling both electron spin waves and magnons. Interconversion among light helicity, electron spin waves and magnons is also discussed. Finally, we show several challenges and provide an outlook on the key steps that must be demonstrated for implementing spin-based wave-parallel computing.