Bitcell With Magnetic Switching Elements For Memory

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current memory technologies face challenges in achieving fast, non-volatile, and dense storage solutions, with magnetic hard disk drives offering low cost but slow access times, while alternative technologies require stable power and have limited life cycles, and existing MRAM approaches are inefficient due to shared read- and write-paths and suboptimal materials.

Innovation Solution

A bitcell design incorporating a magnetic switching cell for data storage and a buffer magnetic switching cell for bitline values, utilizing a write bitline to transfer data via predetermined voltage levels and current direction, and employing magnetic tunnel junctions and resistive devices based on spin- or spin-orbit torque principles to manage resistance states for efficient read and write operations, independent of CMOS devices.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If magnetic hard disk drives are used for storage, then cost per bit is reduced and storage density is increased, but access time becomes excessively long

Engineering Contradiction:
Improvestorage densityVSAvoidaccess time
Core Design Contradiction:
Quantity of substanceVSSpeed

Solution Approach 1:

The patent segments the memory system into distinct magnetic switching cells arranged in an array with separate read and write paths. Each cell is independently addressable through wordlines and bitlines, enabling parallel access operations that reduce access time while maintaining high density through the segmented array structure.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces magnetic switching cells as intermediary elements between the read and write paths. These cells use magnetic domain wall motion as an intermediary mechanism to transfer and store data, enabling fast non-volatile storage without requiring mechanical movement, thus resolving the speed-density tradeoff.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Speed

If alternative memory technologies are used for faster access, then access time is reduced, but storage density decreases and power consumption increases

Engineering Contradiction:
Improveaccess timeVSAvoidstorage density
Core Design Contradiction:
SpeedVSQuantity of substance

Solution Approach 1:

The patent replaces mechanical systems (moving disks in HDDs) with magnetic domain wall motion in solid-state memory cells. This substitution enables fast access times comparable to electronic memory while maintaining the non-volatile and high-density characteristics of magnetic storage, without requiring stable power supply.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Device complexity

If shared read- and write-paths are used in MRAM, then device complexity is reduced, but manufacturing precision and material requirements become suboptimal

Engineering Contradiction:
Improvenumber of transistors per cellVSAvoidmaterial quality
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The patent segments the memory cell into separate read and write paths, each optimized for its specific function. The write path uses magnetic tunnel junctions with specific material stacks for writing, while the read path uses different configurations for sensing, allowing each path to use optimal materials and structures without compromise.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by using different material compositions and structures in different parts of the memory cell. The write path employs magnetic tunnel junctions with specific barrier layers for efficient writing, while the read path uses configurations optimized for low-noise sensing, allowing each region to have the quality needed for its specific operation.

Inventive Principle:
Principle #3Local quality

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 design enables high-density, fast, and reliable non-volatile memory with isolated read and write paths, reducing power consumption and increasing memory density, while being immune to noise sources, allowing for multi-port operations and efficient data storage with minimal sneak current issues.

Implementation Method 1

utilizing a write bitline to transfer data via predetermined voltage levels and current direction, and employing magnetic tunnel junctions and resistive devices based on spin- or spin-orbit torque principles

Methodology Applied
Scientific EffectSpin-transfer torque:

Implementation Method 2

employing magnetic tunnel junctions and resistive devices based on spin- or spin-orbit torque principles

Methodology Applied
Scientific EffectSpin-orbit torque:

Implementation Method 3

an evaluation path whose resistance can switch between a first resistance state and a second resistance state, the first resistance state being lower than the second resistance state, in response to signals magnetically coupled from the programming path

Methodology Applied
Scientific EffectMagnetoresistance: Magnetoresistance

Data Source

PatentUS9524767B2Bitcell wth magnetic switching elements
Publication Date: 2016.12.20 CARNEGIE MELLON UNIV
  • US9524767B2 patent drawing
  • US9524767B2 patent drawing
  • US9524767B2 patent drawing

AI summary

A method includes receiving a data bit value at a buffer in a bitcell based on a first state of a write bitline connected to the buffer, and transferring the data bit value from the buffer to a first magnetic switching cell in the bitcell for a later read operation at least by holding the write bitline to a reference value different from the first state, and asserting first and second predetermined voltage levels on respective first and second write wordlines connected to the buffer.