SSD 신뢰성 향상을 위한 3D 낸드 플래시 메모리 오류율에 따른 K-means 클러스터링 분석 및 차별화된 보호 정책
K-means clustering analysis and differential protection policy according to 3D NAND flash memory error rate to improve SSD reliability
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
3D-NAND 플래시 메모리는 평면적 구조인 2D-NAND 셀을 적층하는 방식으로 단위 면적당 고용량을 제공한다. 하지만 적층 공정의 특성상 각 레이어별 또는 물리적인 셀 위치에 따라 오류 발생빈도가 달라질 수 있는 문제가 있다. 이와 같은 현상은 플래시 메모리의 쓰기/지우기(P/E) 횟수가증가할수록 두드러진다. SSD와 같은 대부분의 플래시 기반 저장장치는 오류 교정을 위하여 ECC를사용한다. 이 방법은 모든 플래시 메모리 페이지에 대하여 고정된 데이터 보호 강도를 제공하므로물리적 위치에 따라 오류 발생률이 각기 다르게 나타나는 3D NAND 플래시 메모리에서는 한계를보인다. 따라서 본 논문에서는 오류 발생률 차이를 보이는 페이지와 레이어를 K-means 머신러닝알고리즘을 통해 군집으로 분류하고, 각 군집마다 차별화된 데이터 보호강도를 적용한다. 본 논문에서는 페이지와 레이어별로 오류 발생률이 현저하게 달라지는 내구성 테스트가 끝난 시점에서 측정된 오류 발생 횟수를 바탕으로 페이지와 레이어를 분류하고 오류에 취약한 영역에 대해서는 스트라이프에 패리티 데이터를 추가하여 차별화된 데이터 보호 강도 제공을 예시로 보인다. 본 논문에서는 기존의 ECC 또는 RAID 방식의 데이터 보호 구조와 비교하여 제안하는 차별화된 데이터보호정책이 3D NAND 플래시 메모리의 신뢰성과 수명향상에 기여할 수 있음을 보인다.
Abstract
3D-NAND flash memory provides high capacity per unit area by stacking 2D-NAND cells having a planar structure. However, due to the nature of the lamination process, there is a problem that the frequency of error occurrence may vary depending on each layer or physical cell location. This phenomenon becomes more pronounced as the number of write/erase(P/E) operations of the flash memory increases. Most flash-based storage devices such as SSDs use ECC for error correction. Since this method provides a fixed strength of data protection for all flash memory pages, it has limitations in 3D NAND flash memory, where the error rate varies depending on the physical location. Therefore, in this paper, pages and layers with different error rates are classified into clusters through the K-means machine learning algorithm, and differentiated data protection strength is applied to each cluster. We classify pages and layers based on the number of errors measured after endurance test, where the error rate varies significantly for each page and layer, and add parity data to stripes for areas vulnerable to errors to provides differentiate data protection strength. We show the possibility that this differentiated data protection policy can contribute to the improvement of reliability and lifespan of 3D NAND flash memory compared to the protection techniques using RAID-like or ECC alone.
참고문헌 (20)
- Cristian Zambelli, "Reliability challenges in 3D NAND Flash memories", IEEE 11th International Memory Workshop(IMW) (2019)
- Rino Micheloni, "Array Architectures for 3-D NAND Flash Memories", Proceedings of the IEEE, 105(9), pp.1634~1649 (2017), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- H. Kim, "Evolution of NAND Flash Memory: From 2D to 3D as a Storage Market Leader", IEEE International Memory Workshop (IMW), pp.1~4 (2017)
- Seung woo Son, "Improve reliability of SSD through cluster analysis based on error rate of 3D-NAND flash and application of differentiated protection policy", 64th Proceedings of KSCI Conference 2021 (2021)
- Agrawal, N., "Design tradeoffs for SSD performance", 2008USENIX ATC, pp.57~70 (2008), USENIX Association
- Seong, Y.J, "Hydra: A Block-Mapped Parallel Flash Memory Solid-State Disk Architecture", IEEE Trans. Comput., 59, pp.905~921 (2010)
- Chen, F., "Understanding Intrinsic Characteristics and System Implications of Flash Memory Based Solid State Drives", Proceedings of the Eleventh International Joint Conference on Measurement and Modeling of Computer Systems, pp.181~192 (2009)
- Jung, M., "An Evaluation of Different Page Allocation Strategies on High-speed SSDs", Proceedings of the USENIX Workshop on Hot Topics in Storage and File Systems (HotStorage) (2012)
- Lee, J, "CPS-SIM:Configurable and accurate clock precision solid state drive simulator", Proceedings of the 2009 ACM Symposium on Applied Computing, pp.318~325 (2009)
- Jaeho Kim, "Improving SSD Reliability with RAID via Elastic Striping and Anywhere Parity", 2013 43rd Annual IEEE, 2013, ~
- Kim, J, "A space-efficient flash translation layer for CompactFlash systems", IEEE Trans. Consum. Electron., 48, pp.366~375 (2002)
- Jaeho Kim, "Building Reliable Massive Capacity SSDs through a Flash Aware RAID-Like Protection", IEEE International Conference on Consumer Electronics (ICCE) 2020 (2020)
- "Flash Memory" (2019)[URL]
- Yuqian Pan, "Process-variation Effects on 3D TLC Flash Reliability:Characterization and Mitigation Scheme", 2020 IEEE 20th International Conference on Software Quality, Reliability and Security (QRS)
- L. Shi, "Exploiting process variation for write performance improvement on nand flash memory storage systems", IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 24, pp.334~337 (2015)
- Yejia Di, "Minimizing Retention Induced Refresh Through Exploiting Process Variation of Flash Memory", IEEE Transactions on Computers, 68, pp.83~98 (2018)
- C. Zambelli, "Characterization of TLC 3D-NAND Flash Endurance through Machine Learning for LDPC Code Rate Optimization", IEEE International Memory Workshop (IMW), pp.1~4 (2017)
- J. MacQueen, "Some methods for classification and analysis of multi- variate observations", Proc. of Berkeley Symp. on Math. Statistics and Prob
- Empirical Measurements of Disk Failure Rates and Error Rates, "Gary, J.", Microsoft, ~, pp.Technical Report MSR-TR-2005-166 (2005)
- Nikolaos Papandreou, "Characterization and Analysis of Bit Errors in 3D TLC NAND Flash Memory", 2019 IEEE International Reliability Physics Symposium (IRPS) (2019)