NAND Memory Internal Copy for Program Fail Recovery

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing memory systems face challenges in efficiently recovering data from NAND program failures, particularly in high-performance SSDs that lack sufficient SRAM/DRAM capacity or do not support XOR techniques for error correction, leading to potential data loss during write operations.

Innovation Solution

Implementing an internal NAND data move operation that allows for data recovery from a cache or temporary location to a new block within the NAND memory, using internal data movement commands and a fire and forget policy to minimize performance impact, and supporting various NAND types including TLC and QLC.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional error correction methods (ECC, XOR parity) are used to handle NAND program failures, then data reliability is improved, but device complexity and resource requirements increase

Engineering Contradiction:
Improvedata reliabilityVSAvoiddevice complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent uses internal NAND copy commands to replicate data from the first portion to the second portion of the nonvolatile memory, creating redundant copies for fault tolerance without requiring external error correction hardware or complex algorithms

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The memory device performs self-recovery by detecting program failures and automatically executing internal data movement operations to restore data integrity, eliminating the need for external controller intervention or complex error correction systems

Inventive Principle:
Principle #25Self-service

2Reliability

If sufficient SRAM/DRAM capacity is allocated for data recovery operations, then data integrity is improved, but cost and device complexity increase

Engineering Contradiction:
Improvedata integrityVSAvoidSRAM/DRAM capacity
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The system uses internal NAND memory copy operations to transfer data between memory portions, eliminating the need for separate SRAM/DRAM buffers by utilizing the nonvolatile memory itself as both source and destination for recovery operations

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The nonvolatile memory serves multiple functions: it acts as both the primary storage medium and the recovery buffer, allowing the same memory cells to be used for data recovery without requiring dedicated SRAM/DRAM resources

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Productivity

If internal data move operations are implemented for failure recovery, then write performance is maintained, but the complexity of memory management increases

Engineering Contradiction:
Improvewrite performanceVSAvoidmemory management complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The memory system automatically detects program failures and triggers recovery operations without host intervention, maintaining transparent performance by handling all recovery operations internally through self-service mechanisms

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system performs data verification immediately after program operations and proactively executes recovery operations before data corruption occurs, preventing performance degradation by addressing failures in real-time rather than during subsequent read operations

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3503107B1Internal copy to handle NAND program fail
Publication Date: 2022.04.06 INTEL CORP
  • EP3503107B1 patent drawingFigure 1~2
  • EP3503107B1 patent drawingFigure 3A~3B
  • EP3503107B1 patent drawingFigure 3C~4

AI summary

An embodiment of a semiconductor package apparatus may include technology to attempt to program data in a first portion of a nonvolatile memory, determine if the attempt was successful, and recover the data to a second portion of the nonvolatile memory with an internal data move operation if the attempt is determined to be not successful. Other embodiments are disclosed and claimed.