Adaptive Memory Codewords Using Mode Sub-Codewords
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data encoding and decoding methods in memory systems face inefficiencies due to the need to align codewords to specific units, leading to unused bits and increased memory initialization time, especially when initializing memory locations without actual data storage.
Innovation Solution
The implementation of a dual-subcodeword system where one subcodeword contains data and parity bits for normal mode and another subcodeword stores mode information, allowing for reduced codeword size in special mode, enabling faster memory initialization by using alignment bits for mode information and reducing the number of bits written to memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If codewords are aligned to specific units with data and parity bits, then error detection and correction capability is improved, but memory initialization time increases due to unused alignment bits
Solution Approach 1:
The patent divides the codeword into two separate sub-codewords: a first sub-codeword containing data and parity bits for error correction, and a second sub-codeword containing mode information. This segmentation allows the mode sub-codeword to be decoded first to determine the operational mode, enabling the data sub-codeword to be processed with the appropriate decoding strength, thereby reducing unnecessary processing time during initialization while maintaining full error correction capability when needed.
2Reliability
If full-strength error correction decoding is applied to all codewords, then reliability is improved, but processing time increases for codewords that do not require error correction
Solution Approach 1:
The patent implements a dynamic decoding approach where the decoding strength is adjusted based on the operational mode indicated by the mode sub-codeword. In normal mode, full-strength error correction decoding is applied to ensure reliability. In special mode (such as initialization), a reduced-strength decoding process is used since error correction is not required, thereby significantly improving processing speed for these operations.
3Manufacturing precision
If alignment bits are used to pad codewords to unit boundaries, then data integrity is improved, but the number of bits written to memory increases
Solution Approach 1:
The patent makes the alignment bits multi-functional by using them to store mode information in the second sub-codeword. This allows the same bits that would traditionally be used for padding and alignment to instead carry operational mode indicators, eliminating the need for separate alignment padding while still maintaining proper data structure and integrity.
Data Source
AI summary
Technology is described herein for encoding and decoding data. A codeword may be divided into two sub-codewords. A first sub-codeword may be used for data bits and associated parity bits. A second sub-codeword may be used for mode sub-codeword. In one embodiment, the alignments bits are used for the mode sub-codeword. The mode sub-codeword stores mode information, in one aspect. The mode information may indicate how the first sub-codeword should be decoded. The mode information includes a normal mode and a special mode. The format and size of the first sub-codeword may vary depending on the mode. The first sub-codeword may be substantially smaller for the special mode. This can be used to vastly reduce the time needed to initialize memory.