Solid State Storage Reading Control via Dynamic Voltage Shift

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional solid state storage devices face inefficiencies in reading data due to the distribution of threshold voltages in NAND-based flash memory cells, leading to erroneous judgments and reduced throughput, especially after repeated program and read cycles that shift the median threshold voltage.

Innovation Solution

Implementing a controlling circuit that performs both hard and soft decoding processes, using multiple sensing voltages to accurately determine cell states and updating these voltages based on histogram parameters and voltage shift functions to improve error correction capabilities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional hard decoding with single sensing voltage is used, then device complexity is low, but measurement precision deteriorates due to threshold voltage distribution and shifting

Engineering Contradiction:
Improvereading accuracyVSAvoiddecoding complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The decoding process is segmented into two distinct stages: hard decoding using a first sensing voltage to obtain initial hard data, and soft decoding using a second sensing voltage to obtain soft data with reliability information. This segmentation allows each decoding stage to use optimized sensing voltages and methods tailored to its specific requirements, improving overall reading accuracy without requiring a complete redesign of the decoding architecture.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces dynamic voltage adjustment by applying different sensing voltages (first sensing voltage for hard decoding, second sensing voltage for soft decoding) based on the decoding stage and requirements. The sensing voltage is dynamically selected and adjusted according to the specific reading conditions, threshold voltage distribution characteristics, and decoding needs, enabling adaptive optimization of measurement precision.

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If multiple sensing voltages and soft decoding are implemented, then measurement precision improves, but loss of time increases due to additional decoding steps

Engineering Contradiction:
Improvereading accuracyVSAvoiddecoding time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Hard decoding is performed as a preliminary action before soft decoding. The hard decoding stage using the first sensing voltage quickly processes the data to obtain hard data and identify potential error patterns. This preliminary processing reduces the burden on the subsequent soft decoding stage, allowing soft decoding to focus on correcting the identified errors rather than processing all data from scratch, thereby reducing overall decoding time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements a feedback mechanism where the results from hard decoding (hard data and error patterns) are fed into the soft decoding process. The soft decoder uses this feedback information along with soft data obtained from the second sensing voltage to more efficiently correct errors. This feedback loop prevents redundant processing and optimizes the soft decoding time by providing targeted correction guidance.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If histogram parameters and voltage shift functions are used, then measurement precision improves through voltage compensation, but device complexity increases

Engineering Contradiction:
Improvethreshold voltage compensation accuracyVSAvoidcontrol circuit complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The control circuit performs self-characterization by automatically generating and storing histogram parameters and voltage shift functions during manufacturing or initial operation. The system uses its own reading operations to collect threshold voltage distribution data, build histograms, and derive compensation functions without requiring external calibration equipment or manual intervention. This self-service approach simplifies the overall system complexity while achieving precise voltage compensation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

Histogram parameters and voltage shift functions are generated and stored in advance during manufacturing or initial setup. This preliminary characterization work is performed once rather than during every reading operation, allowing the control circuit to use pre-computed compensation data to quickly adjust sensing voltages during normal operation, reducing real-time processing complexity while maintaining high measurement precision.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS9490024B1Solid state storage device and reading control method thereof
Publication Date: 2016.11.08 SOLID STATE STORAGE TECH CORP
  • US9490024B1 patent drawing
  • US9490024B1 patent drawing
  • US9490024B1 patent drawing

AI summary

A reading control method for a solid state storage device includes following steps. If a hard decoding process fails, a first histogram parameter and a second histogram parameter are generated in the hard decoding process according to a first sensing voltage, a second sensing voltage and a third sensing voltage. Then, a voltage shift amount is obtained according to the first histogram parameter, the second histogram parameter and a voltage shift function. The first sensing voltage, the second sensing voltage and the third sensing voltage are updated according to the voltage shift amount. Then, a soft decoding process is performed. The updated first sensing voltage, the updated second sensing voltage and the updated third sensing voltage are provided to a non-volatile memory, so that the non-volatile memory generates a soft data.