Reliability characterization and modeling of solid-state drives

Abstract

Solid-state Drive (SSD) technology and product reliability is discussed in this paper from the perspective of overall system reliability and memory subsystem reliability. As relatively new but strongly growing segment of the highperformance storage industry, characterization of SSD reliability may be approached and clarified both from empirical system-level evaluations as well as on the basis of fundamental metrics of wear in the memory subsystem. Further, while expectations of solid-state product reliability may be viewed from a design standpoint of component and assembly, they require reconciliation with empirical observations of field reliability and test validations. The authors illustrate these various approaches based on reliability characterizations and modeling of SSD products. Based on such characterization data, understanding is developed on the impact of operational and environmental factors, such as, wear in the memory subsystem and temperature impact on product reliability. Reliability modeling based on intrinsic error rates of the memory subsystem across SSD populations is also presented to provide statistical estimates of SSD life based on goodness of fit analysis. Thereby, the authors explore and illustrate new conceptual underpinnings by focusing on specific characterizations and modeling towards comprehensive reliability evaluation of SSD technology and products.

A New Post-Fault Reconfiguration Strategy under Open-Phase Operation Conditions of Asymmetrical Double-Star Induction Machines

“…machines because of its ability to reuse conventional three-phase converters [3,4]. Several control strategies have been studied when a fault appears at the power converter or machine level, including the control [4][5][6]. For continuity of service in the case of an OPF, a reconfiguration of the command is mandatory to obtain acceptable performance [7][8][9][10][11][12][13], various measures are necessary (estimation of new reference values for x-y currents [7,8] and reconfiguration of x-y controllers [9,10]), and different control objectives are chosen (minimization of power losses [11,12] and variable current injection [13]).…”

“…The quality of energy delivered by these machines and their tolerance to defects make them more interesting than their three-phase counterparts in applications of comfort, discretion, and mandatory use [1][2][3][4]. Fault tolerance of the multiphase machine was first undertaken in [5] and in recent research to ensure degraded operation without additional external equipment after one or more failures (as long as the number of healthy phases remains greater than or equal to three) and to improve the reliability of the system at the expense of a reduction in power provided after default. DSIM is one of the most used multiphase machines because of its ability to reuse conventional three-phase converters [3,4].…”