Spintronics, which manipulates the spin of electrons rather than just their charge, underpins the development of ultra-low power memory, neuromorphic processors, and stochastic computing. Traditionally, reversing magnetization required a heavy current to inject spins into a magnet, with much of that spin dissipating as waste.
The new study reveals that this loss is not wasted but instead drives spontaneous magnetization switching. Experiments showed that the higher the spin dissipation, the less power was needed to reorient the magnetization, leading to an efficiency improvement of up to threefold compared with conventional approaches.
Crucially, this advance requires no exotic materials or elaborate device structures. The team used a simple device design compatible with standard semiconductor processes, making the technology highly scalable for mass production. This opens pathways to compact, low-power chips for AI, edge computing, and advanced memory.
"Until now, the field of spintronics has focused only on reducing spin losses, but we have presented a new direction by using the losses as energy to induce magnetization switching," said Dr. Dong-Soo Han, senior researcher at KIST. "We plan to actively develop ultra-small and low-power AI semiconductor devices, as they can serve as the basis for ultra-low-power computing technologies that are essential in the AI era."
The researchers expect the principle to accelerate the development of efficient computing systems across multiple domains, including AI accelerators, neuromorphic architectures, and probability-based computation.
Research Report:Magnetization switching driven by magnonic spin dissipation
Related Links
Korea Institute of Science and Technology (KIST)
Computer Chip Architecture, Technology and Manufacture
Nano Technology News From SpaceMart.com
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |
