Magnetization Vector Rotation Reservoir Computing Operated by Redox Mechanism

Abstract

Physical reservoir computing is a promising way to develop efficient artificial intelligence using physical devices exhibiting nonlinear dynamics. Although magnetic materials have advantages in miniaturization, the need for a magnetic field and large electric current results in high electric power consumption and a complex device structure. To resolve these issues, we propose a redox-based physical reservoir utilizing the planar Hall effect and anisotropic magnetoresistance, which are phenomena described by different nonlinear functions of the magnetization vector that do not need a magnetic field to be applied. The expressive power of this reservoir based on a compact all-solid-state redox transistor is higher than the previous physical reservoir. The normalized mean square error of the reservoir on a second-order nonlinear equation task was 1.69 × 10 −3 , which is lower than that of a memristor array (3.13 × 10 −3 ) even though the number of reservoir nodes was fewer than half that of the memristor array.

Fast physical reservoir computing, achieved with nonlinear interfered spin waves

“…Although there was no systematic change with different durations, NMSE dropped significantly at durations of 5 and 20 ns, and the lowest NMSE of 1.39 × 10 −3 was achieved at H of 169 mT and a duration of 5 ns. This value is lower than or comparable to the values for other physical reservoirs that have been reported, in which the NMSEs of experimental physical reservoirs utilizing 90 metal-oxide memristors and with magnetization vector rotation manipulation were 3.13 × 10 −3 and 1.69 × 10 −3 [7,37], and a theoretical physical reservoir utilizing 24 spin torque oscillators was 1.31 × 10 −3 [12]. The NARMA is a more difficult task than the former task since, to predict the output of a NARMA model, a physical reservoir is required not only for nonlinearity but also for short-term memory.…”

“…In the time series data analysis tasks, the readout network of the nonlinear interfered spin wave multi-detection reservoir was trained by ridge regression [30,31,36,37,46]. The reservoir output y(k) shown in equation ( 6) is transformed to;…”

“…After optimizing the W, the computational performance of the physical reservoir was evaluated by calculating 'NMSE' for the second-order nonlinear dynamics equation task and 'NMSE var. ' for the NARMA2 task to compare its performance with that of other physical reservoirs as following equations [7,10,12,13,28,30,31,36,37,46],…”

“…Time-series data processing task (e.g., second-order nonlinear equation task and NARMA2 task) is widely performed to evaluate the computational performance of a physical reservoir [7,10,12,13,28,30,31,[36][37][38][39]46]. The process flow for such tasks is shown in figure 4.…”