2012 · 132 citations · 26 references
ComputingComputer EngineeringComputer ScienceFault ToleranceFault-tolerant ComputingStorage SystemsData DurabilityData Storage SystemsReliable Storage SystemsSolid-state StorageStorage ReliabilityStorage ScalabilityFlash Memory ReliabilityNand Flash MemoryNon-volatile Memory DesignStorage Performance OptimizationStorage System ReliabilitySystem Design OptimizationWrite LatencyWrite Latency Optimization
NAND Flash scaling reduces raw reliability, leading to longer write latencies and complex ECCs that degrade SSD performance, while the required 1‑10‑year retention guarantees far exceed the hours‑to‑days data overwrite patterns seen in datacenter workloads. The study aims to improve SSD performance by exploiting retention relaxation, enabling data to be written at varied latencies or protected by different ECC codes while maintaining reliability. The authors propose a system design that permits data to be written at different latencies or protected by varying ECC codes without compromising reliability. Simulation results show that write speed optimization yields a 1.8–5.7× reduction in response time, and retention relaxation offers both performance and cost advantages for future SSD ECC architectures.
As NAND Flash technology continues to scale down and more bits are stored in a cell, the raw reliability of NAND Flash memories degrades inevitably. To meet the retention capability required for a reliable storage system, we see a trend of longer write latency and more complex ECCs employed in an SSD storage system. These greatly impact the performance of future SSDs. In this paper, we present the first work to improve SSD performance via retention relaxation. NAND Flash is typically required to retain data for 1 to 10 years according to industrial standards. However, we observe that many data are over-written in hours or days in several popular workloads in datacenters. The gap between the specification guarantee and actual programs' needs can be exploited to improve write speed or ECCs' cost and performance. To exploit this opportunity, we propose a system design that allows data to be written in various latencies or protected by different ECC codes without hampering reliability. Simulation results show that via write speed optimization, we can achieve 1:8-5:7× write response time speedup. We also show that for future SSDs, retention relaxation can bring both performance and cost benefits to the ECC architecture.
26
Low-density parity-check codes
Robert G. Gallager · IEEE Transactions on Information Theory · 1962 · 10.6K citations · Full text
On a class of error correcting binary group codes
R. C. Bose, D.K. Ray-Chaudhuri · Information and Control · 1960 · 1.3K citations
Design tradeoffs for SSD performance
Nitin Agrawal, Vijayan Prabhakaran, Ted Wobber et al. · 2008 · 886 citations
Dushyanth Narayanan, Austin Donnelly, Antony Rowstron · ACM Transactions on Storage · 2008 · 759 citations
Failure trends in a large disk drive population
Eduardo Pinheiro, Wolf-Dietrich Weber, Luiz André Barroso · 2007 · 709 citations