Controllability of Relaxation Behavior in Ag-based Diffusive Memristors.
DRC(2023)
摘要
Diffusive memristors - based on the formation and rupture of Ag or Cu metallic filaments within a switching matrix - display a variety of interesting features such as thresholding, self-relaxation, and tunable characteristics of the switching times that can be exploited for applications in emerging memory and neuromorphic computing (NC) areas [1]–[2]. For example, the modulation of the switching times, especially the relaxation time (
$t_{\mathrm{r}}$
) makes diffusive memristors a good fit as a simple and scalable circuit element for temporal- and/or rate-based coding in emerging computing [3]. Despite this potential, a reliable means to modulate
$t_{\mathrm{r}}$
has yet to be found. Here we show a way to control
$t_{\mathrm{r}}$
by employing series resistors in combination with different programming schemes. Additionally, we show a decay of
$t_{\mathrm{r}}$
as a function of programming voltage (
$V_{\mathrm{p}}$
) as opposed to previous observations of rising
$t_{\mathrm{r}}$
with increasing
$V_{\mathrm{p}}$
.
更多查看译文
关键词
Ag-based diffusive memristors,neuromorphic computing areas,programming voltage,rate-based coding,relaxation behavior,simple circuit element,switching matrix
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要