Dual-Layer Data Line Architecture for Memory Throughput
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional memory devices face challenges in balancing data throughput and power consumption, with some devices prioritizing one over the other, making it difficult to design a memory device that optimally combines both improved data throughput and reduced power consumption.
Innovation Solution
The memory device employs a dual-layer data line architecture with top and bottom data lines, allowing for concurrent access and operation on multiple memory cells, coupled with an efficient sense circuit and control mechanism to optimize read and write operations, thereby enhancing data throughput while minimizing power usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional single-layer data line architecture is used, then device complexity is reduced, but data throughput is limited
Solution Approach 1:
The patent introduces a third dimension by stacking data lines vertically across multiple conductive layers. Specifically, first data lines are formed in a first conductive layer while second data lines are formed in a second conductive layer above it, enabling three-dimensional integration. This dimensional transition allows concurrent access to multiple memory cells through different layers, thereby increasing data throughput without proportionally increasing planar area or device complexity.
Solution Approach 2:
The patent segments the data line function across multiple conductive layers, with each layer containing specific data lines that access different memory cell regions. The first data lines in the first conductive layer and second data lines in the second conductive layer operate semi-independently, allowing parallel data transmission channels. This segmentation enables higher throughput by dividing the data access function across multiple spatial segments.
2Productivity
If more data lines are added to increase data throughput, then power consumption increases
Solution Approach 1:
The patent merges multiple data line functions into a stacked architecture where first and second data lines occupy different vertical layers but serve complementary access functions. By combining these layers into a single integrated memory structure, the patent achieves higher throughput without linearly increasing power consumption, as shared control circuitry and overlapping signal paths reduce redundant energy expenditure.
Solution Approach 2:
The patent employs periodic switching between different data line sets in the stacked architecture. During read operations, alternating between first data lines and second data lines allows time-multiplexed access patterns that reduce simultaneous switching activity. This periodic activation reduces peak power consumption while maintaining high average throughput through efficient operation scheduling.
Data Source
AI summary
Some embodiments include apparatuses and methods having a first set of conductive lines, a second set of conductive lines, and memory cells located in different levels of the apparatuses and arranged in memory cell strings. At least a portion of the first set of conductive lines is configured as a first set of data lines. At least a portion of the second set of conductive lines is configured as a second set of data lines. Each of the memory strings is coupled to a respective conductive line in the first set of conductive lines and a respective conductive line in the second set of conductive lines. Other embodiments including additional apparatuses and methods are described.


