Topological magnetic textures for spintronics via machine learning

Magnetic materials are crucial for spintronics, an information technology paradigm in which the magnetic materials´ properties and the electron spin degree of freedom are employed to create new types of devices.

Spintronics offers the prospect ultra-low power devices, through manipulation of spins, without any dissipative charge currents. The emerging field of topological spintronics is particularly promising. Here, two-dimensional skyrmions have been in focus, due to their attractive properties as information carriers.

However, an even more interesting topological soliton exists: the three-dimensional magnetic hopfion.

Magnetic hopfions are a type of three-dimensional magnetic structure characterized by a knotted or twisted spin configuration in a magnetic material.

Since  hopfions can propagate in three dimensions, they open up new avenues in the development of magnetic storage, data processing systems, or neuromorphic devices, and enables logic operations based on braiding.

In this project, the aim is to further the field of  topological spintronics by harvesting and investigating the properties of hitherto unknown magnetic hopfion textures, using an in-house developed machine-learning based method orders of magnitudes faster than alternative methods.

This project will establish innovative solutions for a more sustainabe ICT, and educate the next generation of researchers and experts in topological spintronics, enabling them to tackle key technological challenges.