Graded porous oxide memristor shaped like melting hole

Abstract

The invention provides a graded porous oxide memristor shaped like a melting hole, and the basic structure of the memristor sequentially comprises a bottom electrode layer, oxide layers and a top electrode layer from bottom to top. The memristor is characterized in that the oxide layers serve as resistive function layers and comprise a first oxide layer and a second oxide layer, and the second oxide layer is close to one side of the bottom electrode layer; the second oxide layer is of a porous structure and is of a similar-melting-hole graded porous structure, wherein the porous channels of the porous structure are communicated with one another, and the transportation of metal ions can be regulated and controlled under the action of an electric field; when the external electric field is removed, metal ions can be stored. The oxide memristor has stable, multi-resistance-state and low-energy-consumption memristor characteristics, a new path is provided for exploration of a memristor mechanism, and the electrical performance of the memristor can be improved; the memristor has good characteristics in the aspect of artificial synapse simulation, and a great application prospect is provided for neuromorphic calculation.

Description

technical field [0001] The invention relates to a hierarchical porous oxide memristor based on a cave shape, and the oxide memristor has stable, multi-resistance states, and low energy consumption memristor characteristics, and belongs to the technical field of brain-like computing devices. Background technique [0002] With the continuous improvement of people's requirements for computing performance, a new type of computer with brain-like learning and memory functions has become a research hotspot. The emergence of artificial intelligence has brought great innovations to the development of emerging technologies. Artificial intelligence is a discipline that studies the use of computers to simulate certain human thinking processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.). Among them, learning and application based on neural network is an important way to realize artificial intelligence. Due to its advantages of non-volatility, simpl...

AI Technical Summary

Problems solved by technology

However, memristors still face many problems in the application, such as the resistance switching mechanism is not clear, and the performance is difficult to control, etc.

For example: based on the conductive filament type memristor, due to the random generation and breakage of the conductive filament, the stability of the memristor shows great fluctuations, which limits its application in artificial synapses; and Non-conductive filament-type memristors are at the expense of low switching ratio and poor low-resistance state retention time, which limits their further applications in neuromorphic computing.

Images