Non-Volatile Memory Look-Ahead Reads Without Pre-Read Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The shrinking size of memory circuitry in non-volatile semiconductor memory devices leads to neighbor word line interference (NWI), causing Vth distribution widening and increased read time due to redundant pre-reads, which are inefficient and introduce read disturb.
Innovation Solution
A method and apparatus that skip pre-reads between reads on selected word lines, applying data retention compensation schemes based on identified zones for each grouping of data states, reducing redundant operations and minimizing read disturb.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pre-read operations are performed on adjacent word lines to compensate for data retention loss, then read accuracy is improved, but read time increases and read disturb is introduced due to redundant operations
Solution Approach 1:
The patent extracts and eliminates the redundant pre-read operations from the read sequence. By analyzing which pre-reads are actually necessary and removing the unnecessary ones, the invention reduces read time while maintaining read accuracy through selective application of pre-read operations only where data retention compensation is genuinely needed.
Solution Approach 2:
Instead of performing pre-read operations on all adjacent word lines uniformly, the patent applies partial action by selectively performing pre-reads only on specific word lines where data retention loss is expected to significantly impact read accuracy. This targeted approach avoids the excessive action of universal pre-reading, reducing overall read time while maintaining necessary read accuracy.
2Measurement precision
If pre-read operations are performed on adjacent word lines to compensate for data retention loss, then read accuracy is improved, but read disturb increases due to redundant operations
Solution Approach 1:
The patent removes redundant pre-read operations that contribute to read disturb without providing necessary data retention compensation. By extracting only the essential pre-read operations and eliminating the redundant ones, the invention reduces read disturb while preserving read accuracy where it matters.
Solution Approach 2:
The patent applies partial action by performing pre-read operations only on word lines where data retention compensation is actually needed, rather than uniformly across all adjacent word lines. This selective approach minimizes read disturb by avoiding unnecessary read operations on word lines where they are not required for accuracy.
3Quantity of substance
If memory circuitry is shrunk to increase density, then storage capacity is improved, but neighbor word line interference increases causing Vth distribution widening
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different word lines based on their specific characteristics and proximity to selected word lines. Instead of uniform pre-read operations, the invention selectively applies pre-reads to specific adjacent word lines where neighbor word line interference is most significant, thereby locally addressing Vth distribution widening issues without affecting the entire memory array.
Solution Approach 2:
The patent changes the operational parameters by selectively applying different read voltages and pre-read operations to different word lines based on their proximity and interference characteristics. This parameter variation allows the system to compensate for Vth distribution widening caused by shrinkage-induced neighbor word line interference while maintaining overall storage capacity.
Data Source
AI summary
A memory apparatus and method of operation are provided. The apparatus includes memory cells connected to word lines. The memory cells are disposed in strings and are configured to retain a threshold voltage corresponding to one of a plurality of data states. A control means determines data states for memory cells of at least one word line of the word lines in a pre-read. The at least one word line is adjacent to a selected word line. The control means performs reads on the selected word line for groupings of the plurality of data states in a read operation. Each of the reads including applying a data retention compensation scheme corresponding to a zone identified for a selected set of memory cells of the selected word line. The control means skips the pre-read in between each of the reads on the selected word line for each of the groupings.