Curved FeFET Channel Structure for Linear Synaptic Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current neuromorphic processors face challenges in achieving efficient neural network operations due to non-linear state changes in ferroelectric memory, which affect the accuracy and power consumption of synaptic connections.
Innovation Solution
The design incorporates ferroelectric field effect transistors and capacitors with curved pillar shapes and varying curvatures, allowing for linear state change characteristics and reduced power consumption by optimizing the ferroelectric film thickness and electrode materials, such as TiN and ferroelectric oxides, to enhance synaptic connection strength and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional ferroelectric memory structures are used in neuromorphic processors, then the basic storage function is achieved, but non-linear state changes occur affecting accuracy and power consumption
Solution Approach 1:
The patent applies curvature by designing the channel with curved cross-sections having different curvatures (elliptical pillar shape with continuously changing radius, tapered shape). This curved geometry modifies the electric field distribution within the ferroelectric film, enabling linear state changes that improve synaptic connection accuracy while maintaining energy efficiency in neuromorphic operations
Solution Approach 2:
The patent changes the geometric parameter of the channel from conventional straight/cylindrical shapes to curved shapes with varying cross-sectional areas. This parameter change (curvature) fundamentally alters the electrostatic field distribution, transforming the non-linear switching characteristics into linear state changes, thereby resolving the contradiction between accuracy and power consumption
2Measurement precision
If the ferroelectric film thickness is optimized for linear response, then measurement precision improves, but manufacturing complexity increases
Solution Approach 1:
The curved channel structure (elliptical or tapered) is designed to inherently produce linear response characteristics in the ferroelectric film. The curvature geometry itself serves as the mechanism to achieve linearity, eliminating the need for additional complex control circuits or multi-layer structures, thus improving precision without proportionally increasing device complexity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables more accurate and energy-efficient neural network operations by ensuring linear changes in synaptic connection strength, improving the performance and efficiency of neuromorphic processors in data processing and recognition tasks.
Implementation Method 1
a ferroelectric film surrounding an outer circumferential surface of the channel
Data Source
AI summary
Provided are a ferroelectric field effect transistor, a neural network apparatus, and an electronic apparatus. The ferroelectric field effect transistor includes: a substrate; a source protruding from an upper surface of the substrate in a first direction; a drain protruding from the upper surface of the substrate in the first direction; a channel spaced apart from the upper surface of the substrate and extending between the source and the drain in a second direction different from the first direction; a ferroelectric film surrounding an outer circumferential surface of the channel; and a gate electrode surrounding the ferroelectric film, wherein the channel has curved cross-sections having a plurality of different radii of curvature.