Phys.org September 16, 2024
Magnetic skyrmions are promising candidates for reservoir computing systems due to their enhanced stability, non-linear interactions and low-power manipulation. Traditional spin-based reservoir computing has been limited to quasi-static detection or real-world data must be rescaled to the intrinsic timescale of the reservoir. An international team of researchers (Germany, The Netherlands, Norway) addressed this challenge by time-multiplexed skyrmion reservoir computing, that allowed for aligning the reservoir’s intrinsic timescales to real-world temporal patterns. Using millisecond-scale hand gestures recorded with Range-Doppler radar, they fed voltage excitations directly into their device and detected the skyrmion trajectory evolution. This method was scaled down to the nanometer level and demonstrated competitive or superior performance compared to energy-intensive software-based neural networks. According to the researchers their hardware approach’s key advantage was its ability to integrate sensor data in real-time without temporal rescaling, enabling numerous applications… read more. Open Access TECHNICAL ARTICLE
Schematic representation of the dynamic Brownian reservoir computer. Credit: Nature Communications volume 15, Article number: 8103 (2024)Â