No abstract

With the appealing storage-density advantage, multilevel-per-cell (MLC) NAND Flash memory that stores more than 1 bit in each memory cell now largely dominates the global Flash memory market. However, due to the inherent smaller noise margin, the MLC NAND Flash memory is more subject to various device/circuit variability and noise, particularly as the industry is pushing the limit of technology scaling and a more aggressive use of MLC storage. Cell-to-cell interference has been well recognized as a major noise source responsible for raw-memory-storage reliability degradation. Leveraging the fact that cell-to-cell interference is a deterministic data-dependent process and can be mathematically described with a simple formula, we present two simple yet effective data-processing techniques that can well tolerate significant cell-to-cell interference at the system level. These two techniques essentially originate from two signal-processing techniques being widely used in digital communication systems to compensate communication-channel intersymbol interference. The effectiveness of these two techniques have been well demonstrated through computer simulations and analysis under an information theoretical framework, and the involved design tradeoffs are discussed in detail.

Abstract-Multiple reads of the same Flash memory cell with distinct word-line voltages provide enhanced precision for LDPC decoding. In this paper, the word-line voltages are optimized by maximizing the mutual information (MI) of the quantized channel. The enhanced precision from a few additional reads allows frame error rate (FER) performance to approach that of full-precision soft information and enables an LDPC code to significantly outperform a BCH code.A constant-ratio constraint provides a significant simplification in the optimization with no noticeable loss in performance.For a well-designed LDPC code, the quantization that maximizes the mutual information also minimizes the FER in our simulations. However, for an example LDPC code with a high error floor caused by small absorbing sets, the MMI quantization does not provide the lowest frame error rate. The best quantization in this case introduces more erasures than would be optimal for the channel MI in order to mitigate the absorbing sets of the poorly designed code.The paper also identifies a trade-off in LDPC code design when decoding is performed with multiple precision levels; the best code at one level of precision will typically not be the best code at a different level of precision.

No abstract

No abstract