Sever trace analysis for an optimal design of solid state drives

Research output: Contribution to journalArticlepeer-review

Abstract

Solid state drives (SSDs) that consist of multiple NAND flash memory chips, controller, and an internal firmware are now replacing hard disk drives in server storage markets. Server workloads may have different characteristics from mobile devices and PCs, and thus for a better performance we need to analyze their workload characteristics so that the internal firmware of SSDs utilizes them. In this work, we analyze the representative nine server workloads in the aspects of storage utilization, read request ratio, average request size, inter-arrival time, and sector hot/coldness and present a proper firmware design to utilize each workload characteristics.

Original languageEnglish
Pages (from-to)9696-9699
Number of pages4
JournalInternational Journal of Applied Engineering Research
Volume11
Issue number19
StatePublished - 2016

Keywords

  • NAND flash memory
  • Server workloads
  • Solid state drives

Fingerprint

Dive into the research topics of 'Sever trace analysis for an optimal design of solid state drives'. Together they form a unique fingerprint.

Cite this