
Modern electronics face critical challenges, including high energy consumption and increasing design complexity. In this context, magnonics - the use of magnons, or quantized spin waves in magnetic materials - offers a promising alternative. Magnons enable efficient data transport and processing with minimal energy loss. With the growing demand for innovative computing solutions, ranging from 5G and upcoming 6G networks to neuromorphic computing (mimicking functions of the brain), magnonics represents a paradigm shift that redefines how devices are designed and operated. Developing an innovative magnonic processor that enables highly adaptive and energy-efficient computing was a challenge that Andrii Chumak of the University of Vienna’s Nanomagnetism and Magnonics Group and his collaborators successfully met.
Success Through Trial and Error

Noura Zenbaa, first author of the study, together with her colleagues around Dieter Süss, Physics of Functional Materials at the University of Vienna, built a unique experimental setup using 49 individually controlled current loops on an yttrium-iron-garnet (YIG) film. These loops created tunable magnetic fields to control and manipulate magnons. Using an "inverse-design" approach, the team allowed algorithms to determine the optimal configurations to achieve desired device functionalities, significantly streamlining the design process. After more than two years of development and testing, the team overcame many challenges. "It was a tough journey but seeing it all come together with our first successful measurement was incredibly rewarding", said Noura Zenbaa.
Creating Greener Technologies

Noura Zenbaa, Claas Albert, Fabian Majcen, Michael Kerber, Rostyslav O. Serha, Sebastian Knauer, Qi Wang, Thomas Schrefl, Dieter Suess, Andrii V. Chumak. A universal inverse-design magnonic device.
DOI 10.1038/s41928-024-01333-7

Pictures
Fig. 1: Experimental inverse-design setup consisting of the device itself, mounted between the pole shoes of an electromagnet, vector network analyzer (VNA), the five multi-channel current sources, other components and a PC for running the algorithms required to solve the inverse problem. C: Noura Zenbaa, NanoMag, U of Vienna Fig. 2: Experimental inverse-design setup. Component 1 - universal inverse-design device; Component 2 - electromagnet to apply a magnetic field; Component 3 - Teslameter to measure the field; Component 4 - Vector Network Analyzer (VNA) used as a microwave source and detector; Component 6 - multi-channel current sources; and Component 7 - PC to automate the setup and optimization process. C: Noura Zenbaa, NanoMag, U of Vienna Fig. 3: The three first authors of the paper - Noura Zenbaa (on the right), Claas Abert (on the left) and Fabian Majcen (in the middle) at the moment when the universal inverse-design magnonic device was activated to solve its first problem. C: Andrii Chumak, NanoMag, U of Vienna Fig. 4: The experimental setup. The reconfigurable inverse design device showing the 7×7 omega-shaped loop array placed on a magnetic film. It shows the local magnetic field landscape on the film, displayed in red (film top) and blue (film bottom), generated by the current applied to the loops for both current polarities. The setup includes a vector network analyzer (VNA) in combination with a power splitter and two switches. C: Noura Zenbaa, NanoMag, U of Vienna