BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Inferring Behavior of Solid State Drives
DTSTART:20160503T101500
DTEND:20160503T111500
DTSTAMP:20260408T133358Z
UID:43d6c377e998a7c025658a364f6fc452dc3c6b21db2030df043ee2aa
CATEGORIES:Conferences - Seminars
DESCRIPTION:Ryan Rakvic\, Associate Professor in the Electrical and Comput
 er Engineering Department\, U.S. Naval Academy\, Annapolis\, MD\, USA\nMan
 y government and business professionals indicate there is a shortage of co
 mputer security experts. Hence\, there has been a big push towards enhanci
 ng cyber security\, both in terms of government research dollars and compa
 ny investments. However\, there still exists many unknowns in today's secu
 rity domain\, and one may argue that the problem is worsening. As an examp
 le\, Solid State Drives (SSDs)\, which are becoming popular alternatives t
 o hard disk drives in many systems\, remain a mystery internally. One rece
 nt study claims to prove that not all SSDs may be performing all of their 
 advertised features. TRIM\, as an example\, is a feature that is used by t
 he operating system to notify the SSD that a particular storage location c
 an be erased. However\, the SSD makes no guarantee what it will do with th
 is information. This talk will describe measurement and analysis technique
 s for inferring the behavior of an SSD by observing its power usage. Energ
 y models are derived for three SSDs\, and empirically obtained signatures 
 are presented to identify when a TRIM command is issued and executed. Our 
 findings show that energy models differ among drive manufacturers and they
  also support the hypothesis that not all SSDs may be as advertised. These
  energy measurement techniques could also help future research determine 
 “what else” is happening in the SSDs that we use today.
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR