연구동향 분석
박사

Applying reinforcement learning to mitigate long tail latency problem in SSD

강원경 2020년

논문상세정보
인용/피인용
' Applying reinforcement learning to mitigate long tail latency problem in SSD' 의 주제별 논문영향력
논문영향력 선정 방법
논문영향력 요약
주제
  • 응용 물리
  • Solid state drive
  • garbage collection
  • long tail latency
  • nand flash memory
  • reinforcement learning
  • storage
동일주제 총논문수 논문피인용 총횟수 주제별 논문영향력의 평균
735 0

0.0%

' Applying reinforcement learning to mitigate long tail latency problem in SSD' 의 참고문헌

  • ¡°EnablingCost-effective data processing with smart ssd ,
    pp . 1–12
  • [55] “A persistent key-value store.”
  • [51] J. Do, Y.-S. Kee, J. M. Patel, C. Park, K. Park, and D. J. DeWitt, “Query processing on smart ssds: Opportunities and challenges,” in Proceedings of the 2013 ACM SIGMOD International Conference on Management of Data, SIGMOD ’13, (New York, NY, USA), pp. 1221–1230, ACM, 2013.
  • [45] AVNET, “Zedboard technical specifications,” 2017.
    [2017]
  • [44] N. Megiddo and D. S. Modha, “Arc: A self-tuning, low overhead replacement cache.,” in FAST, vol. 3, pp. 115–130, 2003.
  • [35] G. Bebis and M. Georgiopoulos, “Feed-forward neural networks,” IEEE Potentials, vol. 13, pp. 27–31, Oct 1994.
  • [33] Filebench, “filebench/filebench,” 2016.
  • [2] C. Kim, J. Cho, W. Jeong, I. Park, H. Park, D. Kim, D. Kang, S. Lee, J. Lee, W. Kim, J. Park, Y. Ahn, J. Lee, J. Lee, S. Kim, H. Yoon, J. Yu, N. Choi, Y. Kwon, N. Kim, H. Jang, J. Park, S. Song, Y. Park, J. Bang, S. Hong, B. Jeong, H. Kim, C. Lee, Y. Min, I. Lee, I. Kim, S. Kim, D. Yoon, K. Kim, Y. Choi, M. Kim, H. Kim, P. Kwak, J. Ihm, D. Byeon, J. Lee, K. Park, and K. Kyung, “11.4 a 512gb 3b/cell 64-stacked wl 3d v-nand flash memory,” in 2017 IEEE International Solid-State Circuits Conference (ISSCC), pp. 202–203, Feb 2017.
  • [1] Samsung Electronics Co., Ltd., “Samsung v-nand technology,” 2014.
  • [14] R. S. Sutton and A. G. Barto, Reinforcement Learning: An Introduction. MIT Press, 1998.
  • [13] B. Beyer,C. Jones, J. Petoff, and N. R. Murphy, Site Reliability Engineering: How Google Runs Production Systems. 2016.
    [2016]
  • Waftl : A workload adaptive flash translation layer with data partition
    pp . 1–12
  • Understanding storage traffic characteristics on enterprise virtual desktop infrastructure
    [2017]
  • Tracetracker : Hardware/software coevaluation for large-scale i/o workload reconstruction
    pp . 87–96
  • Tiny-tail flash : Near-perfect elimination of garbage collection tail latencies in nand ssds
    vol . 13 , pp . 22:1–22:26 , [2017]
  • The tail at store : A revelation from millions of hours of disk and { SSD } deployments
    { FAST } 16 ) , pp . 263–276 [2016]
  • The tail at scale
    vol . 56 , pp . 74–80 , [2013]
  • The samsung 970 evo plus ( 250gb , 1tb ) nvme ssd review : 96-layer 3d nand
  • The samsung 860 qvo ( 1tb , 4tb ) ssd review : First consumer sata qlc
  • The crucial p1 1tb ssd review : The other consumer qlc ssd
  • Smartssd samsung @ firstsamsung
  • Self-optimizing memoryControllers : A reinforcement learning approachin Proceedings of the 35th Annual International Symposium onComputer Architecture
    ISCA ’ 08pp . 39–50 [2008]
  • Reviewing the evolution of the nand flash technology
    vol . 105 , pp . 1609–1633 , [2017]
  • Reinforcement learning-assisted garbageCollection to mitigate long-tail latency in ssd ,
    vol . 16 , pp . 134:1–134:20 , [2017]
  • Reducing file system tail latencies withChopper
    [2015]
  • Real-time garbageCollection for flash-memory storage systems of real-time embedded systems
    vol . 3 , pp . 837–863 , [2004]
  • Real-time flash translation layer for nand flash memory storage systems
    pp . 35–44
  • Q-value prediction for reinforcement learning assisted garbageCollection to reduce long tail latency in ssdIEEE Transactions onComputer-Aided Design of IntegratedCircuits and Systems
    [2019]
  • Predicting the memory bandwidth and optimalCore allocations for multi-threaded applications on large-scale numa machines
    pp . 419–431
  • Playing Atari with Deep Reinforcement Learning
    [2013]
  • P. Badia ,, D. Hassabis , D. Wierstra , and C. Blundell , “ Neural episodic control , ” in Proceedings of the 34th International Conference on Machine Learning ( D. Precup and
    of Proceedings of Machine Learning
  • Opportunistic storage maintenancein Proceedings of the 25th Symposium on Operating Systems Principles , SOSP ’ 15
    [2015]
  • Memory bandwidth management for efficient performance isolation in 150multi-core platforms
    vol . 65 , pp . 562–576
  • Marvell nvme pcie gen3x4 ssd controllers
    [2018]
  • Marvell 88nv11xx ssd controllers
    [2017]
  • Making lru friendly to weak locality workloads : a novel replacement algorithm to improve buffer cache performance
    vol . 54 , pp . 939–952
  • Lazyrtgc : A real-time lazy garbage collection mechanism with jointly optimizing average and worst performance for nand flash memory storage systems
    vol . 20 , pp . 43:1–43:32 ,
  • Introduction to neural networks with Java
    [2008]
  • Iii.3 - theory of the backpropagation neural networkin Neural Networks for Perception ( H. Wechsler
    pp . 65 – 93 [1992]
  • I/o trace data files
    [2008]
  • Flashsim : A simulator for nand flash-based solid-state drivesin 2009 First International Conference on Advances in System Simulation
    pp . 125–131
  • Efficient and intelligent garbage collection policy for nand flash-based consumer electronics
    vol . 59 , pp . 538–543
  • Dynamic management of key states for reinforcement learning-assisted garbage collection to reduce long tail latency in ssd
    [2018]
  • Deterministic service guarantees for nand flash using partial block cleaning ,in Proceedings of the 6th IEEE/ACM/IFIP International Conference on Hardware/Software Codesign and System Synthesis , CODES+ISSS ’ 08
    [2008]
  • DFTL : A Flash Translation Layer Employing Demand-based Selective Caching of Page-level Address Mappings
    [2009]
  • D. Yu , N. Choi , N. Kim
    S. Song , Y . Park ,
  • Continuous runahead : Transparent hardware acceleration for memory intensive workloads
    pp . 61:1– 61:12 [2016]
  • Cata : A garbage collection scheme for flash memory file systems
    pp . 103–112 [2006]
  • Cachedgc : Cache-assisted garbage collection in modern solid state drives
    pp . 79– 86
  • Array architectures for 3-d nand flash memories
    vol . 105 , pp . 1634–1649 , [2017]
  • Arpaci-Dusseau ,
  • Actor-critic algorithms , in Advances in Neural Information Processing Systems 12 , S. A. Solla , T. K. Leen , and K. M¨uller , Eds
    pp . 1008–1014 [2000]
  • Accelerating dependent cache misses with an enhanced memory controller
    pp . 444–455
  • ARM (2017), “AI Today, AI Tomorrow : Awareness, acceptance and anticipation of AI : A global consumer perspective”.
  • A cache management scheme for efficient content eviction and replication in cache networks
    vol . 5 , pp . 1692–1701 [2017]
  • , S. Ahn , Y. Hong , I. Yang , B.
    K. Park ,