Announcement

Collapse
No announcement yet.

Please check this BIOS readings

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Please check this BIOS readings

    Hi there!

    I am fighting a really strange memory related problem for quite some time now and I think I really need your help. So I am running on a RIVEBE with 4930K, using a G.Skill 8x8GB Memory Kit with the model number F3-17000CL11Q2-64GBZLD. This machine dual boots OS X 10.11 and Win 8.1, so yes its a so called "hackintosh".

    Now my problem is that I am having Kernel Panics (in OS X), almost always "type 14=page fault" errors. The strange thing is that these most of the time happen only at shutdown / restart (but sometimes also "out of nowhere") and they only happen when I did sleep before. I wasn't able to replicate this in Win though, and also a longer MemTest86 session (about 8 passes in >20 hours) did not reveal any error.

    What was also strange that the problem (almost) went away when I took out 4 out of the 8 DIMMS. And apparently it did not matter which ones, because after swapping these, the problem still wasn't present! Went with 64GB again - and the panics immediatly returned.

    So the question is: what is it? RAM? The IMC? Software? And how can I find out? I already upped VCORE and / or VCCSA, and also upped DIMM voltage a bit to 1.55V but all that did not help. Downclocking RAM to just 800MHz did also not improve the situation.

    Now, whats also strange - at least to me - are some BIOS readings. I saw that some DIMM specific values are not the same across all 4 channels, namely tRWDR, tRWDD, tRWSR and DRAM IOL, these were different between Channels A/C and B/D. Plus I see only three readings of "DRAM CKE minimum pulse width" and not four as I expected.

    Some sign of a problem, or are these values to be expected? Ah and I changed the board as well, going from RIVE to RIVEBE so I am pretty sure the mobo is not the culprit as the symptoms are exactly the same. Plus running at stock settings now, so no OC.

    Any help is greatly appreciated!

  • #2
    Do you have the latest BIOS/EFI?

    What settings did you use? Did you try XMP?

    Memory kit should work flawlessly with that motherboard and CPU. If not, you may need to test them individually to see if any errors are detected. Also, memtest86+ or an updated memory test may be preferred over memtest86

    Comment


    • #3
      Yeah I use the latest BIOS (801).

      Right now, I do not have XMP enabled. I did have before, but downclocked RAM to 1866MHz as I had some errors in IBT then. But the actual problem remained the same. Not sure if it would be stable enough to run that much RAM at relatively high speeds - as I see it this also would put quite some stress on the IMC as well - and while I wouldn't mind having more RAM speed, I would prefer overall stability, and make sure my RAM is absolutely OK in the first place. So when I can be sure that all runs well at stock speed, I consider enabling XMP again.

      And I did use in fact MemTest86+ (the UEFI version) and there, everything was OK.

      So, what do you think of those BIOS readings I described, are they to be expected this way? Maybe thats the first thing to rule out?

      Comment


      • #4
        Manually input the basic DRAM values and see how that works.

        Can you post some pics of BIOS? Mainly CPU and memory settings

        Comment


        • #5
          Yeah sure, here are some BIOS Screens attached as a ZIP File. Thanks again for your help!

          I think this one might be particularly interesting:



          So you can see even the formatting of "DRAM CKE minimum pulse width" just looks like that there is something missing here.

          Also note the different values for tRWDR and tRWDD.

          My question now would be if this is to be expected, just some BIOS display glitch or if in fact this readings could point to a problem with my RAM.

          Note that as I said everything is more or less on stock settings right now, i.e. CPU @x39 and DIMMS @1333MHz so no OC/XMP is effective right now.
          Attached Files

          Comment


          • #6
            Not a big deal either way, but you can manually set them to see what is stable and performs best.

            Comment


            • #7
              OK, so you would say these values are "normal" and at least not a sign of any problems?

              Comment


              • #8
                Correct

                How is the system performance?

                Comment


                • #9
                  System performance seems to be OK. And I finally may have found out what is the actual cause of my problem! Seems to be a memory allocation conflict between a driver in the bootloader and the OS X kernel, so these page faults seem to be software related. At least there is a workaround that allocates more memory and by doing so my problem completely vanished.

                  So I am 99.99% sure, that my RAM seems to have no problem, thanks for helping me clearing this! Seems like I do not need to RMA anything now

                  Comment

                  Working...
                  X