ECC Codeword Distribution Across NVM Dies
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile memory devices experience increased bit error rates due to variance in raw bit error rate (RBER) across a population, leading to non-binomial error count distribution and clustering, which existing error correction codes struggle to manage effectively.
Innovation Solution
Dividing error correcting code (ECC) codewords into portions and storing these portions across multiple dies in a memory device, either physically or logically offset, to achieve a more binomial and tighter RBER distribution, allowing for improved error correction and reduced failure rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If ECC codewords are stored on a single die, then the memory device structure is simple, but the RBER distribution is wide and error correction reliability is poor
Solution Approach 1:
The patent divides each ECC codeword into multiple portions and distributes them across different dies within the memory device. This segmentation of the codeword structure allows the system to achieve a more binomial RBER distribution and improve error correction reliability by reducing the width of the RBER distribution curve, while the overall device structure remains modular and manageable through this systematic division.
2Reliability
If ECC codewords are divided into portions and stored across multiple dies, then the RBER distribution becomes tighter and error correction improves, but the device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing ECC codewords into multiple portions and distributing them across different dies. This reduces the RBER distribution width and lowers the failure rate by making the RBER distribution more binomial, while the complexity is managed through standardized distribution patterns.
Solution Approach 2:
The patent introduces a dimensional aspect by distributing codeword portions across multiple dies in a two-dimensional array structure. This spatial distribution across dies adds a new dimension to error correction, allowing the system to leverage redundancy across multiple physical substrates to achieve tighter RBER distribution.
3Manufacturing precision
If memory device size decreases using advanced lithography, then the manufacturing precision improves, but the bit error rate increases
Solution Approach 1:
The patent segments ECC codewords into multiple portions distributed across different dies. This segmentation provides redundancy that compensates for the increased bit error rates caused by advanced lithography, allowing the system to maintain reliable data storage even as manufacturing precision improvements bring higher RBER.
Solution Approach 2:
The patent implements beforehand cushioning by pre-distributing multiple portions of each codeword across different dies before errors occur. This redundancy structure provides a buffer against the increased bit error rates inherent in smaller, more precision-critical memory cells, ensuring that errors can be corrected even as manufacturing pushes for smaller dimensions.
Data Source
AI summary
Embodiments of the present disclosure describe device, methods, computer-readable media and system configurations for dividing error correcting code (“ECC”) codewords into portions and storing the portions among multiple die of non-volatile memory (“NVM”). For example, a device may include memory including a first die and a second die, and a memory controller configured to store a first portion of a codeword for use with an error correcting code in a first segment of the first die and to store a second portion of the codeword in a first segment of the second die. In various embodiments, the first segment of the second die may be offset from the first segment of the first die. Other embodiments may be described and/or claimed.