Memory Cell Programming for Unbalanced Threshold Voltage Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory devices face performance and lifetime issues due to unbalanced threshold voltage distributions in memory cells, leading to erroneous data sensing when averaging signals across codewords with unequal widths and heights.
Innovation Solution
The solution involves determining the quantity of memory cells to program to each data state within a codeword, allowing for an adjusted average signal to serve as a reference voltage, ensuring accurate data state sensing by balancing the threshold voltage distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If memory cells are programmed to multiple data states to increase memory density, then the memory density is improved, but the threshold voltage distribution becomes unbalanced leading to erroneous data sensing
Solution Approach 1:
The patent adjusts the quantity of memory cells programmed to each data state to balance the threshold voltage distribution. Instead of uniform distribution, it modifies the parameters of cell programming to achieve balanced distributions that prevent sensing errors while maintaining high memory density through multilevel cells.
Solution Approach 2:
The patent employs feedback mechanisms where the system monitors the threshold voltage distribution and adjusts the programming of subsequent cells to maintain balance. This feedback loop ensures that unbalanced distributions are corrected, preventing erroneous sensing while preserving the benefits of multistate memory.
2Ease of operation
If average signal is used as reference voltage for sensing data states, then the sensing process is simplified, but unbalanced threshold voltage distributions cause erroneous data sensing
Solution Approach 1:
The patent modifies the reference voltage selection by choosing specific cell states based on balanced threshold voltage distributions rather than simply averaging all signals. This parameter change maintains the simplicity of using a reference voltage while improving sensing accuracy through strategically selected reference states.
Solution Approach 2:
The patent enables the memory system to self-adjust by selecting reference voltages that inherently account for the balanced distribution. The system serves itself by using the balanced structure to automatically provide accurate reference points without complex external calibration.
Data Source
AI summary
The present disclosure includes apparatuses, methods, and systems for unbalanced programmed data states in memory. An embodiment includes a memory having a group of memory cells, and circuitry configured to determine a quantity of the memory cells of the group to program to a first data state, wherein the determined quantity of memory cells is less than or greater than half of the memory cells of the group, program the determined quantity of the memory cells of the group to the first data state, and program a remaining quantity of the memory cells of the group to a second data state.