Convolutional neural network based on 1T1R and operation method thereof

Abstract

The invention discloses a convolutional neural network based on 1T1R and an operation method thereof. The convolutional neural network comprises an input module, a convolution calculation module, a pooling calculation module and a full connection calculation module. the convolution calculation module, the pooling calculation module and the full-connection calculation module are all composed of 1T1R arrays, synaptic weight information can be stored in situ, fusion of calculation and storage is achieved, consumption of data interaction is omitted, the calculation time is greatly shortened, and meanwhile energy consumption is reduced. Besides, bidirectional gradual change adjustment of the conductance value of the memristor is realized by regulating and controlling the gate voltage of a transistor in the 1T1R device; in this way, in the process of achieving conductivity reduction, the 1T1R device does not need to be operated to the lowest conductivity state again, the target conductivitycan be achieved by directly adjusting the grid voltage of the transistor in the 1T1R device, and power consumption is low.

Description

technical field [0001] The invention belongs to the technical field of artificial neural networks, and more specifically relates to a 1T1R-based convolutional neural network and an operation method thereof. Background technique [0002] With the advent of the era of big data, the demand for computing speed and power consumption of traditional computing systems is increasing, and the von Neumann problem has become an important bottleneck restricting the further development of current computer systems. Since the memristor was proposed, it has become one of the most promising development targets for the next generation of semiconductor memory due to its characteristics of non-volatility, high integration, and low power consumption. The fusion of computing and storage features of memristors has also become a powerful foundation for building new computing architectures. Since the memristor was proposed to simulate the synaptic function of the human brain, the research on synapti...

AI Technical Summary

Problems solved by technology

[0007] Aiming at the defects of the prior art, the purpose of the present invention is to provide a convolutional neural network based on 1T1R, which aims to solve the limitation of the prior art due to its conductance gradient, and the device must be re-operated to the minimum when the conductance is reduced. The problem of high power consumption caused by the conductance state to achieve the target conductance in the conductance-increasing operation mode

Images