Arithmetic Device Magnetic Memory Domain Wall
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current large-scale arithmetic devices, particularly neural network devices, face challenges in achieving both higher scales and energy conservation, which are essential for efficient information processing and reducing power consumption, especially in applications like IoT and AI.
Innovation Solution
The proposed arithmetic device incorporates line-shaped magnetic parts in its memory units, utilizing spin injection for write operations, magnetoresistance for read operations, and magnetic noise generation for energy-efficient operations, allowing for larger scales and reduced power consumption by replacing traditional SRAM elements with magnetic domain wall movement-type recording elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If traditional SRAM elements are used in memory units, then write and read operations can be performed, but chip surface area becomes large and power consumption increases
Solution Approach 1:
The patent replaces traditional electrical-based SRAM elements with magnetic domain wall movement-type recording elements. This substitution uses magnetic field effects instead of electrical circuits, enabling memory operations with significantly reduced chip area and lower power consumption while maintaining write and read functionality through spin injection and magnetoresistance effects
Solution Approach 2:
The patent changes the fundamental operating parameters of memory elements from electrical resistance-based SRAM to magnetic property-based domain wall movement. By utilizing magnetic anisotropy, spin polarization, and magnetoresistance effects, the system achieves the same memory functions with different physical mechanisms that require less space and energy
2Productivity
If the scale of arithmetic device is increased, then information processing capability improves, but power consumption increases
Solution Approach 1:
By replacing traditional electrical memory elements with magnetic domain wall-based elements across the entire arithmetic device, the patent enables large-scale information processing with reduced power consumption per operation, allowing scalability without proportional power increases
Solution Approach 2:
The magnetic domain wall movement-type recording elements serve multiple functions: they provide memory storage, enable logic operations through domain wall manipulation, and facilitate data transfer, replacing multiple separate components that would otherwise be needed in traditional architectures
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 a significant reduction in chip surface area, simplifies operations, and achieves higher energy conservation, making it suitable for large-scale energy-efficient information processing comparable to human brain performance.
Implementation Method 1
utilizing spin injection for write operations
Implementation Method 2
magnetoresistance for read operations
Implementation Method 3
magnetic noise generation for energy-efficient operations
Data Source
AI summary
According to one embodiment, an arithmetic device includes one or a plurality of arithmetic units. One of the one or the arithmetic units includes a memory part including a plurality of memory regions, and an arithmetic part. At least one of the memory regions includes a line-shaped magnetic part.