Intel working on fix for latest SSD firmware issue

Well, all those user reports weren’t flukes. Intel has posted a note on the Support Community forum saying it managed to reproduce a bricking problem induced by its latest solid-state-drive firmware update, and that the company is working on a fix.

Here’s the meaty part from the message by Alan Frost of Intel’s NAND Solutions Group:

We have been contacted by users with SSD issues after using the firmware upgrade tool (version 1.3) in a Windows 7* 64bit environment. Intel has replicated the issue on 34nm SSDs (X25-M) and is working on a fix. If users have downloaded 02HA firmware and not upgraded, Intel recommends they don’t upgrade until further notice. Intel is pursuing the resolution of this as a high priority. No related issues have been reported by users who have successfully upgraded to 02HA firmware via the firmware upgrade tool (version 1.3)."

As we reported last month, Intel pulled the new TRIM firmware for its 34-nm solid-state drives after some users started complaining that the update rendered their drives inoperable. The update normally adds support for Windows 7’s TRIM function, which circumvents the block-rewrite performance penalty by clearing flash pages instead of marking them as available. (Thanks to The Register for the tip.)

Comments closed
    • notfred
    • 10 years ago

    It sounds like the firmware itself is good, just the update tool has some interactions with Windows 7 64bit. I would imagine most of the testing went in to the firmware itself and the how to actually update it was tested less thoroughly.

    I also think that it is great that they asked people to send in drives that they had issues with. There have been some issues that I have been involved in at my work where we have been unable to reproduce the issue in house but the customer can reproduce at will and it turns out to be specific to a particular batch of hardware or to some configuration item that nobody thought was relevant. Intel asking for the drives shows that they were prepared to go to this step when they had difficulties reproducing in house.

    • indeego
    • 10 years ago

    /[<" Intel has replicated the issue on 34nm SSDs (X25-M) and is working on a fix. "<]/ I wonder what attorney approved them posting that message. They basically admitted fault where there was no need to, opening them up to class action and almost certain lawsuits. I'm not saying this isn't bad ethically, but it is *very* bad legally. This is also quite scary technically. This means Intel doesn't really have the best of testing facilities for its products in-house, or anywhere close. You would think the opposite, where they have testing facilities that equal or rival Microsoft's, with thousands of various configs and scripts that automate and verify. Even for a large corp like Intel to semi-beg the community for samples is concerning. Does their storage division do part time work on their integrated graphics by any chance? This is far worse than any Seagate firmware issue, and my technical trust of Intel's storage has ratcheted down several notches. I certainly won't pay premium for these drives anymore. I also can't see Intel fixing this quickly, say before Service Pack 1 for Windows 7. To patch too quickly shows disregard for testing fullyg{<.<}g

      • stdRaichu
      • 10 years ago

      Surely it’s impossible to have truly thorough Q&A in this day and age though? There’s a near infinite variety of hardware/software combinations. I’ve not been following this issue too closely (both my SSD’s are OCZ, all the ones at work are Samsung in laptops or STEC in the servers) but do we know what causes the bricking, or at least various hardware profiles from people who’ve used it? For all I know it might only happen to people using six SSD’s in a bootable RAID6 array on a machine with an IBM chipset using a particular patch level of a RHEL 4 install.

      Surely admitting there’s a fault is a CYA to prevent there being further damage done, Just In Case? If Intel waved their hand and said “there is no fault”, people would install the firmware… and if they bricked their drives you could bet they’d also join the class action, with the promise of bigger payouts since Intel could be shown to be demonstrably negligent. This is damage limitation, not legal suicide.

      Totally agree with the knock to confidence in Intel’s SSD unit though, especially after the way they’ve pooh-poohed people who bought the G1’s. I’m just happy that the Indilinx chaps, after a rather shaky start, appear to have been putting out really solid firmware updates. Fingers crossed anyway 😉

        • kc77
        • 10 years ago

        It’s not impossible at all. There are only four or five south-bridges that I know of that people commonly have and 5 or 6 OS’s that those drives would be commonly found with.

        Considering the issue primarily affects Windows 7 users it’s strange that they didn’t catch it.

        Companies in general are spending less and less when it comes to QA and it’s starting to show.

          • stdRaichu
          • 10 years ago

          Five or six southbridges? Yes. But there are dozens upon dozens of SAS/SATA RAID cards, not to mention each configuration could bring about issues as well. I was just saying it’s impossible (realistically) to thoroughly check anything said HDD can connect to because there are just too many combinations to be had under the storage subsystem.

          In any case, once the fix is released I hope to hell they tell people what the issue was.

      • just brew it!
      • 10 years ago

      I tend not to trust /[

        • indeego
        • 10 years ago

        Intel’s are 13+ months old, and the tech behind SSD has been around for decades, although not in the consumer spaceg{<.<}g

      • Bauxite
      • 10 years ago

      What lawsuits, sure there might be some because someone is always looking for a handout, but they won’t go anywhere meaningful.

      Theres a huge warning before you can flash ‘NOT RESPONSIBLE FOR DATA BLAH BLAH BACK UP YOUR DATA BEFORE IT POOFS’, and you have to type “yes” or “I agree” or some similar thing before it even runs.

      I’m for responsibility as much as the next guy, but no one who ran that didn’t know there were risks and they had to choose to ignore them.

    • OneArmedScissor
    • 10 years ago

    l[

    • MarkD
    • 10 years ago

    I’m told mine is shipping next week from Amazon. We’ll see if I get it. The trim command and firmware upgrade can wait a couple of months.

    • kcarlile
    • 10 years ago

    Still can’t get an X18-M. We ordered a couple a few weeks ago… Wonder if Intel’s pulled the drives again.

    • UberGerbil
    • 10 years ago

    I wonder what all the necessary conditions are to cause this? It has to be pretty obscure; I’m still amazed this got through QA.

    • brm001
    • 10 years ago

    OCZ should be raking it in while this mess is going on–especially for the consumer, their drives are just as good as Intel’s–but I don’t know if it’s working out that way :\

      • adisor19
      • 10 years ago

      Even the OCZ drives have inscribe in price like crazy 🙁

      A month ago, NCIX had the Agility 120GB on special for 314$ CAD with an MIR.

      Now, the same drive is 480$ !!!! i wanna cry :'(

      Adi

      • stdRaichu
      • 10 years ago

      Dunno about the states, but availability for OCZ’s drives has been poor in the UK for months now. I bought my 120GB OCZ Agility for £230 about a week after launch… my usual supplier is now selling them for £330 and they’re back-ordered to hell’n’back. Heck, I pre-ordered my 30GB Vertex for £110, it went on sale at £115 and it’s currently marked at £130 nearly a year after release. Same story for all the other Indilinx SSD’s.

      I think they are reeling it in, at least from the enthusiast sector. Sales must have rocketed after that Anandtech article.

        • VaultDweller
        • 10 years ago

        Availability of Indilinx-based drives in general has been quite poor in Canada, as well.

    • Bauxite
    • 10 years ago

    Demand for these things is still sky high, when they are selling out at hundreds above msrp.

    Trim can wait until they finally bake it proper, it takes awhile to junk up your drive under normal use.

    • 5150
    • 10 years ago

    <insert cartoon here>

Pin It on Pinterest

Share This