Redundant Memory Array Design for Efuse Yield Improvement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing efuse memory technology faces low manufacturing yield due to programming failures and reading errors, which often require large redundancy areas and result in inefficient memory cell replacement.

Innovation Solution

A memory array design with redundant memory cells in each row and column, where each redundant cell stores the same data as the corresponding memory cell, allowing for programming and reading through a specific path formed by MOS transistors and sense amplifiers, enabling data correction and improved yield.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If large redundancy areas are used to compensate for programming failures and reading errors, then manufacturing yield is improved, but device area and complexity increase

Engineering Contradiction:
Improvemanufacturing yieldVSAvoidredundancy area
Core Design Contradiction:
ReliabilityVSArea of stationary object

Solution Approach 1:

The patent merges the functionality of separate redundancy rows and columns into a unified redundant array structure. The redundant memory cells are integrated into the existing array architecture, sharing common bitlines and wordlines with functional memory cells. This consolidation reduces the overall redundancy area compared to traditional approaches that require dedicated redundancy blocks.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The redundant memory cells serve multiple functions: they can replace failed functional memory cells, provide backup storage capacity, and maintain array structural integrity. The same redundant cells can be used for different replacement purposes depending on where failures occur, eliminating the need for separate redundancy structures for different failure modes.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Reliability

If traditional redundancy structures are used to handle programming failures, then reliability is improved, but memory cell replacement efficiency decreases

Engineering Contradiction:
Improveprogramming failure toleranceVSAvoidmemory cell replacement efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the redundancy management into modular units where individual redundant cells can be independently activated to replace failed cells. This segmentation allows for targeted replacement of only the failed memory cells rather than requiring systematic replacement of entire rows or columns, significantly improving replacement efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The redundant memory cells are pre-configured and positioned within the array structure before failures occur. The redundancy mapping and replacement pathways are established in advance, allowing for rapid response to programming failures without requiring complex real-time analysis or extensive reconfiguration operations.

Inventive Principle:
Principle #10Preliminary action

3Ease of manufacture

If efuse melting is used to store data, then manufacturing process simplicity is maintained, but programming precision and reliability decrease

Engineering Contradiction:
Improvemanufacturing process simplicityVSAvoidprogramming precision
Core Design Contradiction:
Ease of manufactureVSManufacturing precision

Solution Approach 1:

The patent implements feedback mechanisms through sense amplifiers that read the resistance state of efuse elements after programming attempts. This feedback allows for verification of successful programming and detection of programming failures, enabling corrective actions such as re-programming or activation of redundant cells, thereby improving overall programming precision without complicating the manufacturing process.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The redundant memory cells are prepared in advance with identical structural and electrical characteristics to functional cells. This preliminary preparation ensures that when replacement is needed, the substitution can be performed with high precision using the same manufacturing processes, maintaining ease of manufacture while improving reliability through the availability of pre-validated backup cells.

Inventive Principle:
Principle #10Preliminary action

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 solution enhances the manufacturing yield of memory devices by allowing accurate data reading and programming even after failures, reducing the need for extensive redundancy areas and improving overall device performance.

Implementation Method 1

The initial resistance of an efuse is considerably small, but when a high current flows through the efuse, the efuse may be melted and the resistance multiplies

Methodology Applied
Scientific EffectJoule heating: Joule Heating

Data Source

PatentUS9418763B2Memory array, memory device, and methods for reading and operating the same
Publication Date: 2016.08.16 SEMICON MFG INT (SHANGHAI) CORP
  • US9418763B2 patent drawing
  • US9418763B2 patent drawing
  • US9418763B2 patent drawing

AI summary

The present invention provides a memory. The memory includes a plurality of memory cells arranged as an array with a plurality of rows and a plurality of column. A memory cell is connected to at least one redundant memory cell in a same row for storing same data as the memory cell; and a column of memory cells correspond to one redundant column of redundant memory cells wherein each redundant memory cell in the redundant column stores same data as the memory cell in a same row.