Abstract
With the shrinking technology size and increasing bits per cell, NAND flash is experiencing increasing number of errors. Error correction code (ECC) is now a key component of flash memory systems. In flash memory, data can be read with different voltages hence with different errors. The read reference voltages are the key factor for RBER seen by ECC. A read failure followed by a read retry happens when errors exceed ECC capacity. Finding the right read voltage with the smallest number of read failures is the key to read performance. The main challenge is that voltages in different cells may drift to different directions by different degrees due to different error sources. Previous methods are designed to either progressively tune the voltage value or empirically predict a read voltage based on error models. This paper proposes a novel approach, by reserving a small set of cells as sentinels, from which the optimal voltage can be inferred, as drifting caused errors exhibit strong locality. Experiments demonstrate the proposed technique is both efficient and effective in finding the optimal read voltage for fast read operations.
Original language | English |
---|---|
Title of host publication | HotStorage'19 |
Subtitle of host publication | Proceedings of the 11th USENIX Conference on Hot Topics in Storage and File Systems |
Publisher | USENIX Association |
Publication status | Published - Jul 2019 |
Event | HotStorage '19: 11th USENIX Workshop on Hot Topics in Storage and File Systems - Renton, United States Duration: 8 Jul 2019 → 9 Jul 2019 https://www.usenix.org/conference/hotstorage19 |
Publication series
Name | HotStorage: Proceedings of the USENIX Conference on Hot Topics in Storage and File Systems |
---|---|
Publisher | USENIX Association |
Conference
Conference | HotStorage '19 |
---|---|
Country/Territory | United States |
City | Renton |
Period | 8/07/19 → 9/07/19 |
Internet address |