Data retention in MLC NAND flash memory: Characterization, optimization, and recovery

TLDR

Retention errors caused by charge leakage over time dominate flash memory errors, and characterizing and reducing them can significantly improve NAND flash reliability and endurance. The study first characterizes how threshold voltage distributions change with retention age in real 2‑nm MLC NAND flash chips and then proposes two new techniques to address these changes. The authors introduce Retention Optimized Reading, which adaptively learns and applies optimal read reference voltages for each block online by periodically estimating a tight upper bound, and Retention Failure Recovery, which recovers data with uncorrectable errors offline by probabilistically correcting cells with retention errors. Characterization reveals that optimal read reference voltages vary with retention age and across memory regions; ROR extends lifetime by 64 % and cuts error‑correction latency by 10.1 % with minimal storage overhead, while RFR halves RBER, effectively doubling error‑correction capability and recovering otherwise uncorrectable data.

Abstract

Retention errors, caused by charge leakage over time, are the dominant source of flash memory errors. Understanding, characterizing, and reducing retention errors can significantly improve NAND flash memory reliability and endurance. In this paper, we first characterize, with real 2y-nm MLC NAND flash chips, how the threshold voltage distribution of flash memory changes with different retention age - the length of time since a flash cell was programmed. We observe from our characterization results that 1) the optimal read reference voltage of a flash cell, using which the data can be read with the lowest raw bit error rate (RBER), systematically changes with its retention age, and 2) different regions of flash memory can have different retention ages, and hence different optimal read reference voltages. Based on our findings, we propose two new techniques. First, Retention Optimized Reading (ROR) adaptively learns and applies the optimal read reference voltage for each flash memory block online. The key idea of ROR is to periodically learn a tight upper bound, and from there approach the optimal read reference voltage. Our evaluations show that ROR can extend flash memory lifetime by 64% and reduce average error correction latency by 10.1%, with only 768 KB storage overhead in flash memory for a 512 GB flash-based SSD. Second, Retention Failure Recovery (RFR) recovers data with uncorrectable errors offline by identifying and probabilistically correcting flash cells with retention errors. Our evaluation shows that RFR reduces RBER by 50%, which essentially doubles the error correction capability, and thus can effectively recover data from otherwise uncorrectable flash errors.