.png){kind=icon}
.png){kind=icon}
Return
Tunable Dynamics via Dual-Ion Modulation for Event-based Data Processing Using a Highly Uniform and Self-Rectifying Memristor Array
.png){kind=icon}
OA
AI
DOI：10.1002/advs.75594.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The demand for new computing architectures has greatly increased with the growth of data and the need for efficient data processing. Memristors are promising candidates for bio-inspired computing hardware owing to their analog behaviors, simple structure, and simple fabrication process. Among them, interface-type memristors have been extensively studied due to their high reliability and low power consumption. Despite these advantages, their spontaneous relaxation properties pose a challenge for bio-inspired computing applications. In this article, we demonstrate the tunable dynamics of an interface-type memristor through dual-ion modulation by incorporating Ag nanoclusters (ion diffusion retarders), and successfully implement a 32 × 32 one-resistor (1R) array (100% yield) with high temporal/spatial uniformity (σ/µ<3%) and high read rectifying ratio (>7 × 104). The Ag nanoclusters combined with the oxygen anions modulate the conductance decay of the memristor by retarding the oxygen anion diffusion, enabling tunable dynamic behaviors. Leveraging these properties, we demonstrate a hardware implementation of the hierarchy of event-based time surface algorithm (HOTS) using a highly uniform interface-type memristor array. By demonstrating successful hardware implementation of exponential decaying kernels in the HOTS algorithm for the classification of moving objects (Poker Dynamic Vision Sensor), this work could pave the way toward resource-efficient computing paradigms based on memristors.
Keywords:
bio-inspired computing
dual-ion modulation
event-based moving object classification
ion diffusion retarder
tunable dynamics
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
14.1
Papers:
1.8W
Citations:
11.5W