Crossbar Memory Encoding for Low-Resistance Current Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In crossbar memory architecture, excessive memristive devices in a low resistive state along a wire segment can lead to excessive electric current, interfering with read/write operations and compromising data signal integrity.
Innovation Solution
A method of encoding data bits to limit the number of crosspoints in a low resistive state to less than half per wire segment, using subsets of crosspoints and encoding schemes like Run Length Limited (RLL) to ensure optimal operating conditions, with cyclical shifts to maximize usable crosspoints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If more memristive devices are placed in a low resistive state to increase data storage capacity, then the storage capacity is improved, but excessive electric current flows through the wire segments causing interference with read/write operations
Solution Approach 1:
The patent applies encoding schemes that transform the representation of data bits by constraining the number of low resistive states along wire segments. By changing the parameter of bit representation from direct mapping to encoded mapping with constraints, the system achieves both high storage capacity and controlled current flow. Specific encoding methods limit the number of '1' bits (low resistive states) to at most half the number of crosspoints per wire segment.
Solution Approach 2:
The patent introduces an encoding layer as an intermediary between the data to be stored and the physical crossbar memory structure. This encoding layer acts as a mediator that translates arbitrary data patterns into constrained patterns that satisfy the current flow limitations while preserving the information. The encoding and decoding processes serve as intermediary transformations that enable the system to overcome the direct contradiction between storage capacity and current control.
2Quantity of substance
If more memristive devices are placed in a low resistive state to increase data storage capacity, then the storage capacity is improved, but noise increases interfering with data signals
Solution Approach 1:
The encoding schemes change the parameter distribution of low resistive states along wire segments by constraining their density. This parameter control ensures that the noise generated by multiple low resistive states remains below threshold levels that would interfere with signal integrity, while still allowing sufficient storage capacity through optimized bit placement and cyclical shifting.
Solution Approach 2:
The encoding layer serves as an intermediary that filters and transforms data patterns to prevent noise-generating configurations. By mediating between the desired high storage capacity and the noise constraints, the encoding/decoding process ensures that only noise-compliant patterns are written to memory, thereby protecting signal integrity while maximizing usable storage.
3Object-generated harmful factors
If encoding schemes are applied to constrain low resistive states, then current control is improved, but device complexity increases
Solution Approach 1:
The patent segments the data writing process into distinct encoding and decoding operations that can be performed independently. The encoding step constrains the data pattern before writing, while the decoding step reverses the constraint after reading. This segmentation allows the complexity to be distributed and managed separately, rather than requiring a single complex operation, thereby making the system more tractable and implementable.
Solution Approach 2:
The encoding process performs preliminary action by pre-constraining the data pattern before it is written to the crossbar memory. This preliminary constraint application prevents the need for complex real-time control during read/write operations. The decoding process then performs the corresponding preliminary action in reverse to recover the original data. By moving the constraint management to preliminary encoding/decoding steps, the operational complexity is reduced.
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 prevents excessive current flow, maintains data signal integrity, and optimizes the use of crosspoints in crossbar memory architecture, enhancing storage efficiency and reducing noise.
Implementation Method 1
A programmable device may be placed at each crosspoint between each wire segment. In one example, crossbar architecture may employ memristors as programmable devices. A memristor is a resistor which is able to change the value of its resistance in response to various programming conditions.
Data Source
AI summary
A method for encoding bits to be stored within a crossbar memory architecture performed by a physical computing system includes designating, with the physical computing system, a subset of crosspoints within a crossbar matrix, the crossbar matrix comprising a number of disjointed intersecting wire segments, the subset corresponding to a predetermined path through the crossbar matrix; and encoding, with the physical computing system, a number of data bits to be placed along the predetermined path; in which the encoding causes bits pertaining to at least one of the wire segments to be subject to a constraint when the data bits are placed along the predetermined path.