High energy-efficient CAM based on single FeFET and its operation method

By adopting a combination of a single FeFET and 2NMOS in the CAM design, combined with adaptive ML precharge and discharge design, the existing CAM design has solved the problems of high energy consumption, large area overhead and insufficient performance, and achieved the effect of energy consumption saving and delay reduction.

CN114678052BActive Publication Date: 2025-03-25ZHEJIANG UNIV
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202210190672.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-02-28
Publication Date
2025-03-25
Estimated Expiration
2042-02-28

AI Technical Summary

Technical Problem

The existing CAM designs have problems with high energy consumption, large area overhead and insufficient performance.

Method used

Using a CAM design based on a single FeFET, each CAM cell consists of 1FeFET and 2NMOS, and reduces energy consumption and delay through an adaptive ML precharge and discharge design.

Benefits of technology

Energy savings and delay reduction are achieved, ML voltage swing is reduced, CAM energy efficiency is improved, and design overhead and production costs are reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114678052B_ABST
    Figure CN114678052B_ABST
Patent Text Reader

Abstract

The present invention discloses a high-energy-efficiency CAM based on a single FeFET and an operation method thereof, and relates to the design of a FeFET-based memory suitable for low power consumption and high performance; the storage characteristics of FeFET are fully utilized to realize the design of a brand-new CAM unit based on a single FeFET, saving the number of transistors, reducing search energy consumption, and obtaining non-volatility of data preservation. The present invention adopts a 2T‑1FeFET structure, combining the advantages of FeFET and CMOS, and without reducing performance, only one FeFET is used to achieve less area overhead and lower energy consumption than the traditional CMOS-based CAM, and non-volatility is achieved; the present invention also adopts a data search design method of adaptive matching line pre-charging and discharging, and further optimizes the search energy consumption by reducing the matching line voltage swing.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Analog content addressable memory utilizing three terminal memory devices

    US20210343341A1