Storage Array Logic Relationship Encoding via Variable Resistance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer storage systems struggle to efficiently store and retrieve logic relationships between objects, which is a fundamental difference from human brain memory, leading to high energy consumption and inefficiencies in data processing.
Innovation Solution
A storage array and chip utilizing a variable-resistance two-terminal device and gating diode connected in series, with a controllable switch between leading-out wires, allowing for bidirectional cut-off and unidirectional conduction states to store and read logic relationship values between objects, enabling efficient storage and retrieval of indirect relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If traditional computer storage systems are used to store logic relationships between objects, then data can be stored at designated addresses, but energy consumption becomes excessively high and retrieval efficiency is poor
Solution Approach 1:
The patent replaces traditional mechanical/electronic storage systems with a neuromorphic storage system that mimics biological neural networks. The storage array uses variable resistance devices to simulate synaptic weights and neural connections, enabling parallel processing of logic relationships between objects. This substitution achieves both low energy consumption (comparable to biological brains at ~20W) and high retrieval efficiency through simultaneous multi-path information processing.
Solution Approach 2:
The patent utilizes variable resistance as a key parameter to encode logic relationship strengths between objects. By dynamically adjusting resistance values in the storage devices, the system can represent different weights of logical connections, enabling efficient storage and retrieval of nuanced relationships. The resistance state changes allow the system to process multiple logic relationships in parallel without proportionally increasing energy consumption.
2Quantity of substance
If data scales are expanded to meet storage demands, then more information can be stored, but it becomes more difficult to acquire useful information from data
Solution Approach 1:
The patent segments the storage system into multiple neural-like processing units that can independently process different logic relationships. Each storage element represents a specific object relationship, allowing the system to handle large data scales by dividing the problem into manageable parallel processing tasks. This segmentation enables efficient information acquisition even as data capacity expands.
Solution Approach 2:
The storage array is designed with universal processing capabilities that can handle various types of logic relationships between objects simultaneously. The same hardware structure can process different kinds of data relationships through parallel activation patterns, making it easy to acquire useful information regardless of data scale expansion. The system universally applies neural processing principles across all stored relationships.
3Reliability
If conventional storage modes are used, then original information data can be stored, but the ability to process and reason about logic relationships between objects is weak
Solution Approach 1:
The patent introduces an intermediary neural processing layer between data storage and retrieval operations. This intermediary layer processes logic relationships by activating neural pathways according to stored weightings, enabling sophisticated reasoning about object relationships while maintaining reliable underlying data storage. The intermediary translates raw stored data into meaningful logical conclusions.
Solution Approach 2:
The storage system performs self-service logic relationship processing through its inherent neural network structure. When queried about relationships between objects, the system automatically activates appropriate neural pathways and computes results through the stored weightings, without requiring external complex processing. The system serves itself by leveraging its own storage structure for processing tasks.
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
The solution enables simultaneous consideration of indirect relationship factors during reading, reducing energy consumption and enhancing computing efficiency, suitable for brain-imitation intelligent applications.
Implementation Method 1
a variable-resistance two-terminal device capable of changing between at least two resistance values under the effect of an electrical pulse signal
Implementation Method 2
the gating diode is forwardly conducted from the first leading-out wire to the second leading-out wire and is reversely cut off from the second leading-out wire to the first leading-out wire
Data Source
AI summary
A storage array and a storage chip and method for storing a logic relationship between objects. The storage array comprises first leading-out wires and second leading-out wires, and a storage unit is connected between each first leading-out wire and each second leading-out wire having different serial numbers. A controllable switch is connected between each first leading-out wire and each second leading-out wire having a same serial number. The storage chip comprises an interface module. A control module is used for producing a control signal. A driving module is used for producing write current, erase current or read current. A first decoder and a second decoder are used for gating the first leading-out wires and the second leading-out wires. A storage array is used for storing a logic relationship value. The storage method comprises write and read operations.