QLC Memory Maintenance Using Stored ECC Data Between Two Passes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Two-pass write operations in quad-level cell (QLC) memory systems can cause damage and reduce reliability due to extended voltage application and increased latency from repeated read operations, leading to uncorrectable errors and reduced memory system performance.

Innovation Solution

Implement a two-portion maintenance operation where the controller stores data from the first error control operation in a portion of the memory system, such as SLC blocks or volatile memory, and uses this data during the second portion to reduce latency and increase reliability by skipping redundant error control operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If two-pass write operations are used in QLC memory systems, then data can be written to multi-level cells, but extended voltage application causes damage and reduces reliability

Engineering Contradiction:
Improvewrite operation capabilityVSAvoidmemory cell reliability
Core Design Contradiction:
Ease of manufactureVSReliability

Solution Approach 1:

The patent segments the write operation into two distinct passes: a first pass that writes data to QLC memory cells using a first voltage distribution, and a second pass that writes the same data using a second voltage distribution. This segmentation allows the system to perform necessary write operations while distributing voltage stress across separate events, thereby reducing cumulative damage to individual memory cells and improving overall reliability.

Inventive Principle:
Principle #1Segmentation

2Reliability

If repeated read operations are performed during two-pass write operations, then data can be verified and corrected, but latency increases due to extended operation time

Engineering Contradiction:
Improvedata integrityVSAvoidoperation latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs error control operations and data verification during the first pass write operation, before the second pass is executed. By preliminarily checking and correcting data in the first pass, the system reduces the need for repeated read operations during the second pass, thereby maintaining data integrity while minimizing additional latency.

Inventive Principle:
Principle #10Preliminary action

3Quantity of substance

If voltage is applied for extended duration to program QLC states, then data can be stored in quad-level cells, but memory cell damage increases

Engineering Contradiction:
Improvedata storage capacityVSAvoidvoltage damage
Core Design Contradiction:
Quantity of substanceVSObject-affected harmful factors

Solution Approach 1:

The patent divides the voltage application process into two separate passes with different voltage distributions. Instead of applying a single extended voltage pulse that would cause cumulative damage, the system applies voltage in two distinct events with intermediate recovery time, allowing memory cells to withstand the total voltage exposure required for QLC programming while reducing peak stress and damage accumulation.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS20260100238A1Multi-level cell maintenance operations
Publication Date: 2026.04.09 MICRON TECHNOLOGY INC
  • US20260100238A1 patent drawing
  • US20260100238A1 patent drawing
  • US20260100238A1 patent drawing

AI summary

Methods, systems, and devices for multi-level cell maintenance operations are described. A controller of a memory device may perform a two-portion maintenance operation in order to transfer data from a first block of memory cells to a second block of memory cells. For example, during a first portion of the maintenance operation, the controller may read first data from the first block of memory cells and perform a first error control operation to correct the first data. The controller may store second data associated with the first error control operation to volatile memory of the memory device in response to performing the first error control operation. During a second portion of the maintenance operation, the controller may perform a second error control operation using the second data associated with the first error control operations stored to the volatile memory of the memory device.