Distributed Precharge Circuit Units for Semiconductor Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional semiconductor memory apparatuses face increased precharge time due to line load issues, leading to deteriorated operational characteristics and reduced speed, as longer local IO lines increase resistance and capacitance, affecting tCCD and tWRT delays.
Innovation Solution
The semiconductor memory apparatus includes precharge circuit units aligned in the extension direction of data input/output lines, spaced apart to minimize line load, and a precharge controller that distributes the precharge signal through multiple paths with varying delay units to ensure simultaneous activation, thereby stabilizing and accelerating the precharge operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the local IO lines are extended to connect more memory banks, then the connectivity and coverage are improved, but the line load increases due to resistance and capacitance, causing precharge time to increase
Solution Approach 1:
The patent divides the single precharge circuit unit into multiple precharge circuit units (first precharge circuit unit and second precharge circuit unit) that are distributed at different positions along the local IO lines. Each precharge circuit unit independently precharges the local IO lines in its vicinity, segmenting the precharge function to reduce the effective precharge distance and line load for each unit, thereby reducing overall precharge time while maintaining coverage of extended memory bank connections
2Device complexity
If a single precharge circuit unit is used, then the device complexity is reduced, but the precharge time increases due to the inherent resistance and capacitance of extended lines
Solution Approach 1:
The patent segments the precharge function by placing multiple precharge circuit units at different positions along the local IO lines. This segmentation allows each unit to handle a smaller portion of the total line length, reducing the RC time constant for each segment and enabling faster overall precharge operation
Solution Approach 2:
The patent introduces spatial distribution as a new dimension by positioning precharge circuit units at different locations along the local IO lines rather than using a single centralized unit. This spatial arrangement allows parallel precharging of different line segments, effectively reducing precharge time without proportionally increasing complexity
3Speed
If multiple precharge circuit units are distributed along the local IO lines, then the precharge time is reduced by minimizing line load, but the device complexity increases
Solution Approach 1:
The precharge function is segmented into multiple distributed units, each handling a specific segment of the local IO lines. This segmentation enables faster precharge operation by reducing the effective line length and RC time constant for each unit, while the modular nature of the segmentation allows for systematic design and implementation
4Device complexity
If the precharge signal is transmitted through a single path, then the control logic is simplified, but the precharge operation becomes unstable due to varying signal delays
Solution Approach 1:
The patent introduces multiple signal transmission paths by distributing precharge circuit units at different positions along the local IO lines. Each unit receives and processes the precharge signal through its own path, allowing simultaneous independent operation that stabilizes the overall precharge operation by eliminating signal delay variations that would occur in a single sequential path
Data Source
AI summary
Disclosed is a semiconductor memory apparatus capable of improving precharge performance. The semiconductor memory apparatus includes a plurality of memory banks, data input/output lines commonly connected to the memory banks, and a plurality of precharge circuit units connected to the data input/output lines and aligned in an extension direction of the data input/output lines while being spaced apart from each other by a predetermined distance.