Announcement

Collapse
No announcement yet.

List of Motherboards with issues when running MemTest86 in multi-CPU selection modes

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • @David,

    I reran the MemTest86 v8.4 Pro build version 1001 in Single CPU mode with the suggested CPU #1 (rather than CPU #0).

    In this attempt, I tried limiting the DIMM count to 1 and I still encountered the hang.

    The Supermicro BIOS version 2.1 was released 5 months ago (in February 2020).

    I’ve attached the new log files with the corresponding screenshot of the last time updated. I am unable to post the other screenshot unless I create a double forum post.
    Attached Files

    Comment


    • Hello,

      I was running memetest v8.4 on my Asus Crosshair VII hero trying to find why I get random freeze. And on the last pass, I got this error:

      [UEFI Firmware Error] Could not start CPU 15

      Not sure I can get some logs. It just started back on the Test 13, pass 4 after a few minutes "sleeping" I think

      Comment


      • You can add me, been tearing my hair out over this and glad to finally know what the problem is.
        • ASUS M5A99FX Pro R2.0
        • 2501
        • AMD FX-9590
        • Crashes PC when attempting to run in Parallel and other multi CPU modes. Test runs perfectly fine in single CPU mode and the initial tests on CPU 1 in the multicpu modes run fine and when it goes to run tests after "own address 1 CPU" crashes consistently. I also crash outside of Memtest86 but under the same circumstances (simultaneous multi core and heavy RAM usage). Tested with all except cores 1 and 2 disabled and always crashes in multicpu tests. But runs fine with two CPUs enabled.

        Comment


        • lilloog, what BIOS firmware version are you running? And can you post the log with the lock up in it. Then we can either blacklist the motherboard and/or ask ASUS to fix it.

          Comment


          • Believe that two of the Supermicro motherboard I'm using are having this issue as well.

            X10SAE and the X10SLH-F

            Both boards have their BIOS updated with the last versions, 3.2 and 3.3a respectively. They both run a Xeon E3 1231 v3 and 32 GB of 18KSF1G72AZ-1G6E1.

            Running the test is parallel causes the system to freeze; however this only occurs after 30 hours of running the test, which I understand is quite long, but the test proceeds fine in single CPU mode and no memory errors occur in either mode, the system just becomes completely unresponsive. I also experience system hangs on other OSes.

            Both file is too big to upload, so I've included a link to download them:

            https://www.dropbox.com/sh/um5816k6z...6pr4j6lga?dl=0

            If it determined that this issue is caused by something else then I will move to another thread.

            Comment


            • If it only fails after 30 hours it sounds more like hardware failure of some sort. (there is more load on the system in multi-threaded mode than single threaded).

              Comment


              • As of today I have replaced every component in the server, RAM was oddly enough the last thing I replaced and after running the test for another 30 hours I once again hanged with the same messages in the log file.

                2021-02-20 05:43:14 - RunMemoryRangeTest - CPU #2 timed out, test time = 0ms (BSP test time = 6273ms)
                2021-02-20 05:43:14 - RunMemoryRangeTest - CPU #4 timed out, test time = 0ms (BSP test time = 6273ms)
                2021-02-20 05:43:14 - RunMemoryRangeTest - CPU #6 timed out, test time = 0ms (BSP test time = 6273ms)

                Repeated through out the log file until it freezes.

                Comment


                • There's also a likely possibility that there is a bug in the UEFI firmware, such as a resource leak related to multiprocessing.

                  It may be worth contacting Supermicro support to look into it (though not sure how much time they are willing to spend to investigate/debug the issue as in the past they haven't been very responsive)

                  Comment

                  Working...
                  X