Boolean logic implementation method, unit and circuit based on storage and calculation integrated transistor

Abstract

The invention discloses a Boolean logic implementation method, unit and circuit based on a storage and calculation integrated transistor. The Boolean logic implementation method is implemented by utilizing characteristics of the storage and calculation integrated transistor and a read-write mode of the storage and calculation integrated transistor. A basic unit is composed of a pull resistor and a storage and calculation integrated transistor, wherein the threshold voltage of the storage and calculation integrated transistor can be regulated by an external physical field; a pull resistor and a transistor in the basic unit are connected in series, and a transistor grid is independent; and the basic unit can realize sixteen Boolean logic operations on the basis of data storage through different voltage configuration modes and simple unit cascading and combination. According to the invention, a plurality of logic operations can be realized by using a smaller number of transistors than traditional CMOS transistors, the circuit design area is greatly optimized, and the problems of power consumption and time delay caused by data transportation between the storage unit and the data unit are effectively solved.

Description

technical field [0001] The invention belongs to the technical field of semiconductors and integrated circuits, and in particular relates to a Boolean logic implementation method, unit and circuit based on a storage-computing integrated transistor. Background technique [0002] In a traditional computing system based on the von Neumann architecture, due to the physical separation of the computing unit and the storage unit, data needs to be frequently transferred between the two, resulting in a serious loss of system power consumption and speed. The performance improvement of the memory is far behind the performance of the CPU, and the performance gap between the two continues to widen, which limits the development of high-speed and low-power processors. At the same time, applications such as high-memory access, high-parallel artificial intelligence, and brain-like computing are also accelerating the development of in-memory computing. [0003] Memory-computing integrated tra...

AI Technical Summary

Problems solved by technology

[0002] In a traditional computing system based on the von Neumann architecture, due to the physical separation of the computing unit and the storage unit, data needs to be frequently transferred between the two, resulting in a serious loss of system power consumption and speed

The performance improvement of the memory is far behind the performance of the CPU, and the performance gap between the two continues to widen, which limits the development of high-speed and low-power processors

Images