During the last couple of weeks, numerous reports of Samsung SSD 840 and 840 EVO having low read performance have surfaced around the Internet. The most extensive one is probably a forum thread over at Overclock.net, which was started about month ago and currently has over 600 replies. For those who are not aware of the issue, there is a bug in the 840 EVO that causes the read performance of old blocks of data to drop dramatically like the HD Tach graph below illustrates. The odd part is that the bug only seems to affect LBAs that have old data (>1 month) associated with them because freshly written data will read at full speed, which also explains why the issue was not discovered until now. 

Source: @p_combe

I just got off the phone with Samsung and the good news is that they are aware of the problem and have presumably found the source of it. The engineers are now working on an updated firmware to fix the bug and as soon as the fix has been validated, the new firmware will be distributed to end-users. Unfortunately there is no ETA for the fix, but obviously it is in Samsung's best interest to provide it as soon as possible.

Update 9/27: Samsung just shed some light on the timeline and the fixed firmware is scheduled to be released to the public on October 15th.

I do not have any further details about the nature of the bug at this point, but we will be getting more details early next week, so stay tuned. It is a good sign that Samsung acknowledges the bug and that a fix is in the works, but for now I would advise against buying the 840 EVO until there is a resolution for the issue.

POST A COMMENT

116 Comments

View All Comments

  • Kristian Vättö - Friday, September 19, 2014 - link

    Technically yes. ECC and parity both have a limit of how much they can recover and if there are too many corrupted bits, then the whole data will be corrupt.

    But yeah, that was just a theory I came up with. Once I have some actual information, I'll be sharing is ASAP.
    Reply
  • Kristian Vättö - Friday, September 19, 2014 - link

    *sharing IT ASAP.

    I guess that's a sign that I need to call it a day and get some sleep.
    Reply
  • hojnikb - Friday, September 19, 2014 - link

    this site REALLY needs an EDIT function :) Reply
  • Coup27 - Friday, September 19, 2014 - link

    Definitely. Reply
  • hrrmph - Saturday, September 20, 2014 - link

    Nope. It's a feature not a flaw.

    And it's working just as it should.

    Good on Kristian for calling it the way it is and admitting that he is too tired to post.
    Reply
  • Solid State Brain - Tuesday, September 23, 2014 - link

    If old data is affected, it either means that:

    1) Wear leveling isn't working properly and static data is not getting periodically reshuffled to new memory cells. The controller likely takes more time to read old static data in order to increase read operation reliability, but under normal usage patterns and no firmware bugs users aren't supposed to experience this. This bug would imply that a portion of NAND will have worn up more over time - bad news, but hopefully fixable.

    2) Wear leveling is working properly, but somehow old data remains tagged as old even after getting reshuffled/refreshed to different NAND flash cell locations. In other words the SSD controller is acting needlessly conservatively in order to avoid read errors.

    To me, the way this issue is showing up points to SSD controller bugs rather than inherent problems with the TLC NAND memory of Samsung 840 drivers (at least, not directly), which means that a firmware update should indeed be able to fix it.
    Reply
  • kgh00007 - Tuesday, September 23, 2014 - link

    If that's the case I won't be buying one of these ever again! MLC all the way from now on, my 250GB Crucial M500 mSata has been out performing my 120GB 840 EVO since I got it.

    If this is true and you don't use a drive for a few months or a year you could loose all your data??
    Reply
  • Coup27 - Friday, September 19, 2014 - link

    Kristian, good work on this being swiftly on the case. However, I am rather puzzled by your recommendation not to buy the drive until a fix has been issued. The EVO has been a solid drive since it's release and recommended many times by every man and his dog. This issue is not a critical one but more of an annoyance which has taken nearly a year to surface. Add in the fact a fix is probably only a few weeks away, your statement does seem rather dramatic. Reply
  • hojnikb - Friday, September 19, 2014 - link

    To be fair, EVO is not exactly the best value for quite some time. So really, if you're in a market for budget drives, there are better/cheaper options, unless you're after nieches (1TB msata for example).

    If lots of people recommend it, it doesn't make it a good value. I mean, SSD market is chaning quite quickly and one day this drive is a great value and the next isn't anymore, since competition is offering more for less.
    Reply
  • Coup27 - Friday, September 19, 2014 - link

    I want to retract my comment. On second reading you simply said to hold off buying the drive until a fix has been released. Reply

Log in

Don't have an account? Sign up now