Solid-state drives lose data if left without power for just a few days

Discussion in 'hardware' started by Minimalist, May 10, 2015.

  1. quietman

    quietman Registered Member

    Joined:
    Dec 27, 2014
    Posts:
    511
    Location:
    Earth .... occasionally
    I fully agree on that point , and I tried to make that clear in my post ..... but if it wasn't clear enough , then the fault was mine .
    Minimalist used text from the original ZDNet article as the title for the thread , and was entirely correct in doing so.

    Bill_Bright wrote :-

    Was the data verified intact before put in storage?
    Was the device properly stored in it's plastic case the entire time?
    Did the handler take the essential ESD precautions before, during and after storage?
    Was the device left undisturbed throughout the many years in storage?
    What were the actual environmental conditions (temperature, humidity, exposure to light) while in storage?


    Oh Yes Bill .... now that is a major factor !

    By their very nature , thumb drives ( USB sticks ) are often in trouser pockets and the whole ESD factor comes into play .
    And there are so many variables here .... Polyester trousers or Natural fibre ( eg Wool ) trousers ....
    ......Leather shoes or synthetic shoes ... walking on carpet or on industrial flooring.

    The Van Der Graaf generator and the Wimshurst machine were invented by combining exactly these elements ....
    ... insulator material , naturally electro-static material .... plus motion .

    The trouser pocket is potentially a mini-version of these devices .
    If anybody doubts this , think of the times when you have touched an elevator door , a water-pipe , or even a car on a hot , dry day
    .... and you got a sudden jolt .... sometimes you can even hear it !

    Imagine what could happen if that ZAP is discharged into the SSD that you just touched , or that's in your pocket !

    I spent many years in the electronics industry , and the various company's policies on ESD protection were often strongly enforced ....
    .... eg . " at all times , employees must be wearing grounded wrist-bands , working on anti-static mats ... etc ."
     
  2. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    I've been an electronics tech since 1972. This was before the effects of ESD on high-density, solid-state, digital devices was understood - mainly because back then, integrated circuits (ICs), weren't very dense at all. The problem with ESD and today's ICs is that a static discharge can be so tiny that we humans cannot see, feel, or hear it, yet the discharge (arc) is more than enough to scorch a Grand Canyon sized gorge (microscopically speaking) through millions of transistor gates on the "chip" without us even being aware a discharge occurred.

    And yeah, pockets are great for that and (besides lint and cookie crumbs) are a primary reason to always use the cap on the device to cover and protect the electrical contacts.
     
  3. Keatah

    Keatah Registered Member

    Joined:
    Jan 13, 2011
    Posts:
    1,029
    What's worse is a small dent in a circuit trace, put there by a weak ESD blast. This is latent damage that shows up later. This is exactly like a pothole. As the electrons flow through it and around it, the hole gets bigger and bigger. First it shows no symptoms, then some intermittent operation, and finally complete failure.

    There is no time schedule by which this progresses. Most of the time the damaged area exhibits increased migration of materials away from the point of interest. Insulators can become conductors or develop resistance. Conductive pathways can open completely, or short against adjacent traces. This can snowball into damaging other nearby elements.
     
    Last edited: May 19, 2015
  4. Gullible Jones

    Gullible Jones Registered Member

    Joined:
    May 16, 2013
    Posts:
    1,466
    A couple thoughts:

    a) USB flash drives are cheaper and slower than SSDs. On the one hand that might imply lower flash memory quality. On the other, SSDs might be built to finer tolerances, and therefore less likely to take kindly to heat, ESD, etc.

    b) I was just reading a book on backup solutions, from back in 2007. It mentions that hard disks - the spinning kind - may have iffy data integrity if disconnected from a power source for too long. (It still recommends them heavily over tape and optical media, though.)

    I'm not really a hardware person though, and haven't been in IT very long, so I'm not really sure what to think.
     
  5. Keatah

    Keatah Registered Member

    Joined:
    Jan 13, 2011
    Posts:
    1,029
    SSD's use the same flash and geometry as USB sticks. There is no significant difference, the chips are the same, but laid out differently for speed. Parallel you know. The problem with SSD today is that the industry beancounters clamored and pushed for higher density chips. And therein lies the problem. Corporate greed and profits pushing a technology past its breaking point.

    You can read this this and this and that and everything. The industry journals publish what they want to make sales or push a product to market or make it obsolete.

    HDD spinners have much better retention rate than flash. And much better than the industry wants you to believe. All you really need to do with Spinners is power them up to help distribute the lubrication. It isn't necessary, but it helps the drive last longer when you actually go and use it.

    In any case I've got a box of disks from 1994 that are still working. I filled them up and every 5 years I take one out and test it. Have about 6 more to go. And these last 6 will have been without power for many many years. Once I power one up, it is no longer "valid" for the test, because, well, it got power applied. But all the files they return pass CRC just fine.
     
  6. luciddream

    luciddream Registered Member

    Joined:
    Mar 22, 2007
    Posts:
    2,545
    I had a box with a Samsung 850 Pro SSD (2.5") turned off for about 2 months now, booted it up and everything was just fine.
     
  7. xxJackxx

    xxJackxx Registered Member

    Joined:
    Oct 23, 2008
    Posts:
    8,625
    Location:
    USA
  8. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    Good find! That says it all.

    Now it will be interesting to see if the original KoreLogic blog by Don Allison that falsely mis-represented the facts and started this information debacle will go back, fess up to his mistake and retract his statement. Or at least simply apologize for not clearly stating the facts. Same with the ZDNet guy who claimed it was "new research".

    I bet not, and instead, if anything, they will attempt to dance around the subject to cover up their incompetence - totally ignoring the well known fact that the cover-up is always worse than the original misdeed.

    Too bad Don Allison closed his blog for comments after he created it or this mess and unnecessary panic may have been averted way back in March. Their email, btw, should anyone want to complain, is info_2015@korelogic.com.

    More good info from PC World: http://www.pcworld.com/article/2921...ally-need-to-worry-about-ssd-reliability.html

    http://www.pcworld.com/article/2856052/grueling-endurance-test-blows-away-ssd-durability-fears.html
     
  9. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    Not really. The original article of 2010 still stands true, even though a bit exaggerated.
    What has changed in the industry of ssds between 2010 and 2015 to make Mr. Cox change his mind?
    The answer is simple in 2010 Seagate did not produce any ssd drive and the exaggeration would help boost the sales of seagates hdd drives, but since 2013 seagate came in the ssd market and that same article could hurt also the sales of seagate's ssd drives.
    It is all about marketing.

    Panagiotis
     
  10. quietman

    quietman Registered Member

    Joined:
    Dec 27, 2014
    Posts:
    511
    Location:
    Earth .... occasionally
    .... my edit above ....

    If I had to make a choice , I'm probably going to put more trust in magnetic storage , if only from instinct , but I seriously doubt that ANYTHING bad
    is going to happen to data on my SSDs , within the next decade , even if they don't see a DC voltage for the whole 10 years !

    This entire thing may be yet another example of certain types of business making big bucks by creating , or increasing paranoia
    among their potential customers....
    ... " it's the modern world " , and in some business sectors , it has become the goose that lays the golden eggs.

    As an example , the companies that make so-called "security" equipment for airports must be in Seventh Heaven ....
    .... more paranoia = more profit .

    Panagiotis said it right .... " It is all about marketing ".
     
  11. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    Nothing!!!! He didn't change his mind! Please read the PCWorld article again because, sorry, but it seems you missed or misunderstood what the PCWorld article stated. :( Mr. Cox clearly noted to PCWorld that readers misunderstood the data he presented back in 2010.

    And sorry, but because you have misinterpreted the PCWorld article, your comments about Alvin Cox, his association with Seagate and marketing hype is all wrong too! When the original article was written by Cox, he wrote it talking as the Chairman of JEDEC and it was published by JEDEC. NOT as a Seagate employee or as a Seagate document.

    Note the JEDEC Member List consists of dozens of IT HD/SSD/memory companies including Western Digital, Corsair, Fujitsu, HGST, Intel, JMicron, Kingston, Micron, Samsung, Toshiba and more.

    So again, because he was talking as the JEDEC Chairman, it was a JEDEC document and NOT a Seagate document!
     
  12. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    Sorry,but you are the one that misinterpreted it.
    And by the way when he wrote the original Jedec article he was not talking as a JEDEC Chairman but as Seagate chairman.;)

    Article.jpg

    Panagiotis
     
  13. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    By the way for those that intend to use ssds as storage devices it would be wise once a year to plug them on a system and perform a read write sequence. The problem is not related with Alvin Cox original article/misinterpretation but to the fact that according to the Jedec standards JESD218 the retention of data requirement for a client class ssd is 52 weeks when stored(powered off) at 30C. What happens after that I guess it depends on the manufacturer

    On the other hand hdds according to a 2008 research http://lockss.org/locksswiki/files/ISandT2008.pdf should have a retention of data for 20-30 years (depends on the humidity) when stored at 30C.

    Panagiotis
     
  14. Keatah

    Keatah Registered Member

    Joined:
    Jan 13, 2011
    Posts:
    1,029
    In case I haven't mentioned it before, I have HDD that are more than 30-years old and they're still retaining data.
     
  15. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    No. You are wrong on both counts.

    Again, read the PC World article. In there Cox explained what the original article was about - yet you deny that. o_O

    And it is a JEDEC article, not a Seagate article - as your own image shows. Yes, he was chairman of both JEDEC and Seagate at the same time, but the article was written as a JEDEC article. You can refuse to accept all the evidence presented in this thread by me and everyone else if you choose, but it still does not change the facts, or make you right.
     
  16. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    Said from someone who posted earlier in the same thread :rolleyes:
    It seems that you confuse your misinterpretations and changes of mood with other people's way of thinking...

    And the Chairman for the JC-64.8 was Frank Chu (HSGT) and Vice-Chair was Andrew Peng (Fujitsu).
    https://www.jedec.org/about-jedec/committee-chairs

    Panagiotis
     
    Last edited: May 30, 2015
  17. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    I did change my position on Alan Cox after I read more and learned all the facts. And that is why I said you need to read what has been posted since the beginning of this thread to see what Alan Cox really meant, and understand how that original article was mis-represented by ZDNet and others as "new" research. It was not new research and, in fact, it did not apply to normal consumers, but heavy users with SSDs that had been used more than normal consumers would ever likely do.

    I have NOT, however, changed my position at all on SSDs.

    :( Not in 2010 when that article was written!
     
  18. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    Bill as I said in my first post in the thread
    ...plus that article was made before the tests of client class ssds were completed...
    In one of the new PC World articles http://www.pcworld.com/article/2925...nt-lose-data-if-left-unplugged-after-all.html they say that
    Could you please find for me, where in the jedec specs/standards for client class ssds is mentioned the fact that the data much be retained for 52 weeks (powered off state, 30C temp.) after the disk ended it's lifetime? o_O
    If that is not marketing I do not know what it is...

    Panagiotis
     
  19. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    How is that marketing? JEDEC is not a marketing agency, nor does it produce anything. It is a consortium of manufactures and developers that sets standards. And it includes SSD as well as HD makers. It is similar to IEEE and W3C.

    As for your 52 week question, pages 26 and 27 but again, that has nothing to do with marketing because again, JEDEC consists of companies from a variety of IT fields.
     
  20. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    I am not talking about the Jedec article (even though it was compiled before the tests finished which is not very scientific), I'm talking about the PC world article and Cox responses (if really were his responses and not misinterpretation from the journalist?).

    And pages 26, 27 say nothing about end of life of the ssds. They only describe the requirements for the drives to be tested and that is that client ssds must retain data for 52 weeks when in use at 40C 8hrs/day or when powered off at 30C.

    Panagiotis
     
  21. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    :( You have to take everything in context - that is, the original article and the PC World article where Alan Cox clarifies it.
     
  22. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    Bill, please reread carefully the original jedec article.
    It describes the requirements of the drives to be tested and the methodology of the tests to be conducted it gives examples, estimations and workloads... but does not give any results of the tests that were conducted. Where are the results of those tests? Gives clarification on what exactly? Where is the statistical data?

    Panagiotis
     
  23. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    No. There's no need. I have read it all several times now and I have the gist of the topic from the original JEDEC article, to the sensationalized, irresponsible misrepresentation of the original article, to the author's clarification of the original article. I know what I need to know to make informed decisions for my future builds, for my client builds, and when giving advice to other seeking it.

    If you want more than the original intent from that 5 year old article, please go and complain to the author.
     
  24. pandlouk

    pandlouk Registered Member

    Joined:
    Jul 15, 2007
    Posts:
    2,976
    Same here. And no, I have no need to complain... others were complaining for several weeks now...
    And articles like that, that talk a lot about how the research is/will be done, but tell nothing at all about the results need zero clarification because there is nothing to clarify.
    As I said earlier is all about marketing, nothing more and nothing less

    Panagiotis
     
  25. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    :argh:

    Come on dude! Marketing has absolutely nothing to do with any of this. Who is trying to market what?

    Not one of those articles, not the original JEDEC, not the bloggers who referenced it, not the several PCWorld articles clarifying it have anything what so ever to do with selling anything - except maybe ZDNet subscriptions. So you can keep saying it, but it just ain't so.

    Corsair makes SSDs but does not make hard drives and they are members of JEDEC.
    Kingston makes SSDs but does make hard drives and they are member of JEDEC.
    PNY makes SSDs but does not make hard drives and they are members of JEDEC.
    ADATA makes SSDs but does not make hard drives and they are members of JEDEC.
    Micron makes SSDs but does not make hard drives and they are members of JEDEC.
    ...and so on and so on.

    Note the one and only reference in the original JEDEC article for "hard" drives says only that SSDs,
    So once and for all, this has nothing to do with marketing.
     
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.