Product-Code ECC Decoding with Stored RS Syndromes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor storage devices with error correction circuits using product code frames face limitations in further improving error correction capability, as they require re-reading and re-generating syndromes for different error correction levels, which is inefficient and time-consuming.
Innovation Solution
The implementation of a semiconductor device with a memory controller that includes separate storage regions for successful and unsuccessful ECC decoding, allowing for iterative correction using RS syndromes without re-reading all ECC frames, enabling ECC decoding at different levels without re-generating syndromes, and storing RS syndromes for computation and rollback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction capability is improved by using a product code frame, then error correction capability is enhanced, but decoding time and processing complexity increase due to requiring re-reading and re-generating syndromes for different error correction levels
Solution Approach 1:
The patent pre-calculates and stores syndromes for all possible error patterns in a lookup table before actual decoding operations. When decoding is needed, the system simply queries the pre-computed syndromes rather than recalculating them, significantly reducing decoding time while maintaining high error correction capability through the product code frame structure
Solution Approach 2:
The patent performs preliminary organization of ECC frames into successful and unsuccessful categories, and pre-computes RS syndromes for iterative correction before actual error correction is needed. This preliminary preparation eliminates the need for time-consuming re-reading and re-generation of syndromes during actual decoding operations
2Reliability
If error correction capability is improved by using a product code frame, then error correction capability is enhanced, but processing complexity increases due to multiple decoding passes and syndrome regeneration
Solution Approach 1:
The patent segments the decoding process into distinct stages: frame collection processing that categorizes ECC frames into successful and unsuccessful groups, and iterative correction processing that handles only the unsuccessful frames. This segmentation reduces processing complexity by avoiding redundant operations on already-corrected frames while maintaining comprehensive error correction capability
Solution Approach 2:
The patent extracts and isolates only the unsuccessful ECC frames for iterative correction processing, separating them from successfully decoded frames. This extraction approach reduces processing complexity by focusing computational resources only on frames that actually require correction, rather than processing all frames through multiple complete passes
3Reliability
If re-reading and re-generating syndromes is performed for different error correction levels, then comprehensive error correction is achieved, but efficiency decreases due to redundant operations
Solution Approach 1:
The patent pre-computes and stores syndromes for different error correction levels in advance, organizing them by error pattern severity. When decoding is required, the system queries these pre-computed syndromes based on the actual error conditions rather than regenerating them, achieving comprehensive error correction across multiple levels while maintaining high decoding efficiency
Solution Approach 2:
The patent discards successfully decoded ECC frames from further processing and focuses iterative correction only on unsuccessful frames. By recovering and re-processing only the necessary subset of frames through multiple error correction levels, the system achieves comprehensive error correction while avoiding redundant operations on already-corrected data
Data Source
AI summary
A semiconductor device of an embodiment includes an ECC decoding processing circuit configured to perform ECC decoding on ECC frame data in a lateral direction of a product code frame, an RS decoding processing circuit configured to perform Reed-Solomon (RS) decoding on second frame data in a longitudinal direction of the product code frame, a memory M0 in which a syndrome generated for the ECC frame data decoded is stored, a memory M1 in which an RS syndrome generated for ECC frame data for which the ECC decoding has been successful is stored, and a memory D in which ECC frame data for which the ECC decoding has been failed is stored as frame data which cannot be corrected through decoding, and frame collection processing, and iterative correction processing of performing RS decoding on the uncorrected frame data collected in the frame collection processing are executed.