Classification Code Decoding for Soft Data in Memory ECC

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing memory error correction techniques face challenges in efficiently utilizing soft data, which is slower and consumes more bandwidth compared to hard data, and require multiple reads to determine reliability, impacting memory throughput and error correction performance.

Innovation Solution

The use of a classification code (CC) decoder to generate soft data based on hard reads from memory, which can be used in conjunction with LDPC codes, allowing for improved bit error rate (RBER) performance with minimal ECC overhead, and enabling efficient operation by leveraging only a small portion of the ECC overhead.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If soft data is determined by a soft read, then error correction performance is improved, but read speed decreases and bandwidth consumption increases

Engineering Contradiction:
Improveerror correction performanceVSAvoidread speed
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The patent segments the data processing into two distinct paths: hard read for speed and soft read for reliability. The hard read provides initial data quickly, while the soft read is selectively applied only to portions of data that require error correction, thus maintaining overall system speed while improving error correction performance where needed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies soft read selectively rather than to all data. By identifying which specific data portions require error correction and applying soft read only to those portions, the system achieves improved error correction performance without the full bandwidth consumption and speed penalty of universal soft read.

Inventive Principle:
Principle #16Partial or excessive action

2Reliability

If soft data is determined by a soft read, then error correction performance is improved, but bandwidth consumption increases

Engineering Contradiction:
Improveerror correction performanceVSAvoidbandwidth consumption
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts only the necessary soft data information needed for error correction rather than processing all data through soft read. By identifying and extracting only the specific data portions that require error correction, the system reduces bandwidth consumption while maintaining improved error correction performance.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies soft read selectively to only those data portions that require error correction, rather than applying it to all data. This partial application reduces the overall bandwidth consumption associated with soft read operations while still achieving improved error correction performance where needed.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If multiple reads are performed to determine reliability, then soft data accuracy is improved, but memory throughput decreases

Engineering Contradiction:
Improvesoft data accuracyVSAvoidmemory throughput
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the reading process into a first read for initial data retrieval and a second selective read for reliability determination. By dividing the reading process and applying the second read only to specific data portions that require error correction, the system maintains high memory throughput while achieving accurate soft data where needed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs multiple reads selectively rather than universally. By identifying which specific data portions require reliability determination and performing additional reads only for those portions, the system achieves accurate soft data for error correction while minimizing the impact on overall memory throughput.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2946386B1Determining soft data using a classification code
Publication Date: 2020.02.05 MICRON TECHNOLOGY INC
  • EP2946386B1 patent drawingFigure 1
  • EP2946386B1 patent drawingFigure 2
  • EP2946386B1 patent drawingFigure 3~4

AI summary

Apparatuses and methods for determining soft data using a classification code are provided. One example apparatus can include a classification code (CC) decoder and an outer code decoder coupled to the CC decoder. The CC decoder is configured to receive a CC codeword. The CC codeword includes a piece of an outer code codeword and corresponding CC parity digits. The CC decoder is configured to determine soft data associated with the piece of the outer code codeword, at least partially, using the corresponding CC digits.