Self-compliance and forming-free memristor arrays with a SiO2 scavenging barrier for energy-efficient neuromorphic computing

초록

In this work, low-power RRAM (Resistive random-access memory) devices were characterized by a SiO2 layer serving as an oxygen scavenging barrier, which suppresses conductive filament overgrowth and reduces operation current and power consumption. Additionally, the integration of an AlOx/TiOy overshoot suppression layer enabled intrinsic self-compliance and forming-free operation. XPS analysis confirmed the oxygen composition of the oxygen-rich TiOy, and it was also verified that the higher oxygen composition in TiOy suppresses filament formation, which decreases the operation current. Consequently, the SiO2 layer decreases the LRS current significantly by 103 times. Furthermore, the device demonstrates the endurance of 6 x 103 cycles and retention over 104 s, maintaining analog multi-bit operation. When the proposed device was integrated into an 8 x 8 passive array and programmed with the designated weights, a low mean absolute error (MAE) of 10.8 nA was achieved. Additionally, vector-matrix multiplication operations demonstrated excellent accuracy, with over 99% of results falling within an error margin of 10%. Based on this low MAE, MNIST image classification simulations were conducted, yielding classification accuracy exceeding 96%.

키워드