Memory Erase Retention Using Pre-Erased Free Pools

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional memory subsystems face challenges in maintaining data reliability due to charge gain issues in non-volatile memory cells, particularly in QLCs and PLCs, leading to reduced read window budget and performance degradation, which are not adequately addressed by existing techniques like erase on demand.

Innovation Solution

Implementing a memory subsystem that immediately erases memory upon invalidation and performs charge gain checks, maintaining a free pool of memory through shallow erase operations to ensure data reliability and improve sequential write throughput.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If erase on demand is used to address charge gain issues, then data reliability is improved, but read window budget is reduced and performance degrades

Engineering Contradiction:
Improvedata reliabilityVSAvoidsequential write throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by performing erase operations in advance on memory blocks that are invalidated or no longer needed, rather than waiting for demand. This allows the memory subsystem to have pre-erased blocks ready for future writes, improving sequential write throughput while maintaining data reliability through proactive charge gain prevention

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements beforehand cushioning by maintaining a pool of pre-erased memory blocks that serve as a buffer for incoming write operations. This cushioning mechanism ensures that when write operations are requested, there are already prepared blocks available, preventing performance degradation while ensuring data reliability through proper erase retention management

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

2Reliability

If immediate erase upon invalidation is implemented, then charge gain issues are prevented, but system complexity increases

Engineering Contradiction:
Improvedata reliabilityVSAvoiderase management complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies self-service by implementing automatic tracking and management of invalidated memory blocks. The memory subsystem autonomously identifies blocks that need erasing, performs the erase operations, and maintains the free pool without requiring complex external management, thereby preventing charge gain issues while managing system complexity through automated processes

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent implements feedback mechanisms where the memory subsystem continuously monitors the state of memory blocks, detects invalidation events, and triggers appropriate erase operations. This feedback loop ensures data reliability through proper charge gain prevention while managing complexity through systematic state tracking and automated response

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS20260064281A1Optimizing data reliability using erase retention
Publication Date: 2026.03.05 MICRON TECHNOLOGY INC
  • US20260064281A1 patent drawing
  • US20260064281A1 patent drawing
  • US20260064281A1 patent drawing

AI summary

Methods, systems, and apparatuses include moving a portion of memory to a garbage pool in response to determining that the portion of memory is invalid. The portion of memory is erased in response to determining that the portion of memory is invalid. A request to move an additional portion of memory to a free pool from the garbage pool is received. A free pool includes a queue including erased portions of memory, which serve as next portions of memory to fulfill subsequent cursor requests. The erased portion of memory is moved from the garbage pool to the free pool.