No announcement yet.

X58A-UD3R rev 1.0 & Xeon X5650 & Mushkin 12gb pc3-12800 - UNSTABLE

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

  • X58A-UD3R rev 1.0 & Xeon X5650 & Mushkin 12gb pc3-12800 - UNSTABLE

    Hi guys,

    I've been using a couple of existing threads on here to try and solve my problems, but sadly to no avail, so here I am making my own post to hopefully get to the bottom of this.

    I've done a reasonable amount of OCing in the past, so i'm fairly savvy with what needs changing etc, but this is infuriating me now,

    I recently got hold of 12gb of mushkin ram for cheap at work. I ran a memtest on all 6 sticks (not individually) for 24 hours before buying and there were no errors at all.

    12gb of Mushkin RAM (product code 998946) PC3-12800 9-9-9-24.

    It's 6 sticks of 2GB (2 sets of matching 3).

    Prior to this RAM, I had 6gb of Corsair XMS3 and it was as stable as a rock, even with a hefty overclock to 4.1 on my Xeon X5650 (Beast of a cpu, best £40 upgrade ever).

    Now, however, i CANNOT get this 12gb mushkin stable. Sometimes it doesn't post properly and resets CMOS to default before continuing, sometimes it gets to windows 8.1 loading and blue screens (memory_management usually), then the next time that incredibly annoying windows repair menu launches (and is useless as none of it seems to work), sometimes even that menu crashes and resets...but here's the thing, sometimes it will just WORK!

    These varying occurences can happen without changing any settings in between at all. Once windows boots up it's solid. No slowdowns, all 12gb recognised by windows and CPU-Z. Gaming for hours, Movie watching for hours, installing and uninstalling, virus scanning, and using PLEX to stream. It was most recently on and solid for 3 days non-stop. Reset it when I got back home to it, and now it's buggered again...NO SETTING CHANGED.

    The first thing I did was remove my CPU OC, as the timings were all pretty high and specific to my previous set of RAM. Optimized defaults tried...crashed at windows load, fail-safe defaults....same, crashed on windows boot

    I then upped my DRAM voltage to 1.66 as per a lot of recommendations, then upped my CPI/VVT to manage the extra ram, and this seems to have a sweet-spot around 1.4 (where I most recently had it stable). I've tried with XMP on, and XMP off. I've tried with Turbo,standard and extreme. I've gone through so many variations of settings that SHOULD work, but nothing seems stable.

    I'm not certain that it isn't a hardware issue, but the fact that when it DOES boot, it performs flawlessly and is stable for days makes me think the RAM is fine, it's just some settings that aren't managing to load windows comfortably.

    It's so tricky to pinpoint the issue for me, as there seems to be no patterns to break down. The blue screen 75% of the time seems to be when the windows desktop should appear once loaded. I've already had to system restore once because this crash (At that vital time during windows loading) screwed with my registry/drivers and gave me this freaky flickering desktop causing recurring hardware issue.

    I'm hoping someone here could shine some light on what 'DOES WORK' ie, don't touch it, it won't cause these issues, what i should be changing, or maybe you can spot something between my hardware that just outright isn't compatible?

    Thanks SO MUCH in advance for your extra knowledge, I've already learn't a lot from lurking here!

    Maybe i'm not destined to use 12gb ram via 6 sticks. Sigh </3

  • #2
    Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

    Wow... ****ty nobody has replied in 6 days.... what happens if you only run one kit?

    When in blue screens what is the error?

    Vin
    Main Rig
    OS = Win10-64Bit
    CPU = Ryzen 1700x Overclocked to 4Ghz with custom water-cooling loop
    Mem = 16GB RAM @ 3200Mhz
    MB = Asus ROG C6H
    GPU = Asus 1080Ti ROG Strix
    HD = 512GB M.2 Samsung 960 Pro
    PSU = EVGA SuperNOVA 1300w
    Case = Cooler Master HAF-X 945

    HTPC / Home Server
    OS = Win7 64Bit running XBMC HTPC Front end with Windows Server 2012 Virtual Machine with 12GB ram assigned for homer server
    CPU = i7-980X @ 3.5Ghz CoolerMasster Hyper 212 Evo
    Mem = 24GB RAM
    MB = Gigabyte GA-X58A-UD5 v1.0 with F6 Bios
    GPU = EVGA nVidia GT210
    HD = 2x PNY 120GB Raid 0 (OS)
    Storage = 8TB WD Black Storage
    PSU = Corsair TX750
    Case = Define R5

    Comment


    • #3
      Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

      Originally posted by Vinster View Post
      Wow... ****ty nobody has replied in 6 days.... what happens if you only run one kit?

      When in blue screens what is the error?

      Vin
      Hi there,

      Sorry for my own slow response, still having PC issues and still haven't 100% solved it!

      I've since ran memtest86+ on sticks individually for at least 7 hours each (10+ passes minimum) and only one had errors. I re-tested this same one a few days later and it went for a whole 24+ hours without an error. VERY confusing?! How could this be? These tests were on a spare PC. My next step is to memtest them on this same PC

      I'm still getting blue-screens on windows launch, but as before, once it's been up and running for a while it's all good, until I cold boot again.

      It's ALWAYS a cold boot where the issues occur.

      I'd like to add i'm not even using 6 sticks at the moments. I'm only using 3. Once i've got these figured out and stable, then i'll advance onto the extra 3 sticks!

      I recently reset CMOS, and at default settings it was blue screening every time....SO i jumped into BIOS, dropped the RAM speed to 1066 (8x SPD), and upped the Voltage to 1.6 (even thought it's rated at 1.5v 9-9-9-24?) and bumped the QPI a little. Turbo enhancement, XMP Disabled, and it then booted straight away without an issue.

      I'm getting a mixture of BSOD codes. Here's a screenshot of my viewer...



      Referencing this - forums.tweaktown.com/gigabyte/43058-bsod-error-codes.html

      It seems it's voltage/setting issues, of the RAM sticks are faulty. If they're faulty then suurely memtest would find errors?

      I'm borderline giving up with this :(

      Is there a way I can show you guys my BIOS settings for you to spot and definite no-no's I may have overlooked?

      Thanks a bunch
      Attached Files

      Comment


      • #4
        Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

        Originally posted by Megatonks View Post
        I'd like to add i'm not even using 6 sticks at the moments. I'm only using 3. Once i've got these figured out and stable, then i'll advance onto the extra 3 sticks!

        I recently reset CMOS, and at default settings it was blue screening every time....SO i jumped into BIOS, dropped the RAM speed to 1066 (8x SPD), and upped the Voltage to 1.6 (even thought it's rated at 1.5v 9-9-9-24?) and bumped the QPI a little. Turbo enhancement, XMP Disabled, and it then booted straight away without an issue.

        I'm getting a mixture of BSOD codes. Here's a screenshot of my viewer...
        BSODs still with only 3 sticks? Are they installed in 3 channel mode?
        Do you still have the old RAM ?

        Comment


        • #5
          Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

          Do you have a picture of where you installed the RAM (slots). I'm wondering if a pin in the slot got damaged? also are all the connection points on the ram sticks OK?

          Looking at those faults, they are memory related. I remember 1A and 3B when testing tight memory timings.

          Vin
          Main Rig
          OS = Win10-64Bit
          CPU = Ryzen 1700x Overclocked to 4Ghz with custom water-cooling loop
          Mem = 16GB RAM @ 3200Mhz
          MB = Asus ROG C6H
          GPU = Asus 1080Ti ROG Strix
          HD = 512GB M.2 Samsung 960 Pro
          PSU = EVGA SuperNOVA 1300w
          Case = Cooler Master HAF-X 945

          HTPC / Home Server
          OS = Win7 64Bit running XBMC HTPC Front end with Windows Server 2012 Virtual Machine with 12GB ram assigned for homer server
          CPU = i7-980X @ 3.5Ghz CoolerMasster Hyper 212 Evo
          Mem = 24GB RAM
          MB = Gigabyte GA-X58A-UD5 v1.0 with F6 Bios
          GPU = EVGA nVidia GT210
          HD = 2x PNY 120GB Raid 0 (OS)
          Storage = 8TB WD Black Storage
          PSU = Corsair TX750
          Case = Define R5

          Comment


          • #6
            Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

            Thanks for the replies guys!

            Yes i've still got the old RAM, (tempted to test again), and yes, i'm getting these errors with just the 3 sticks now. It's a nightmare!

            I just ran a memtest 86+ for 7 hours overnight with 6 passes (6 and a half) and no errors... that was with these 3 sticks in. All find and dandy.

            Will memtest detect RAM errors if they're not directly caused by RAM faults? eg... if BIOS settings aren't suitable and cause errors etc?

            The sticks are installed in 3 channel mode, the white slots, as per the mobo manual :) Slots 1, 3 and 5.

            Booted fine again this morning (although it was on all night doing memtest, so technically not a cold boot).

            Here's a SS of HW Monitor and CPU-z on my pc right now...

            http://i.imgur.com/wZQxp2Z.png

            (won't embed, too large)

            Would damaged pins in the slots be something i'd be able to see do you think? although the old RAM has been fine! hmmm

            Hypothetically, what will I need to change when adding the other 3 sticks to try 12 gb again? Just DRAM Voltage (maybe) and the QPI? I took the QPI fairly high last time and it didn't achieve stability...but maybe there was something else going on.

            Comment


            • #7
              Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

              Originally posted by Megatonks View Post
              Yes i've still got the old RAM, (tempted to test again)
              Yeah that was the idea, just to make sure this still works.
              Suppose you first boot until post/BIOS/boot menu, and then reboot would there be errors? I mean how long should it be between cold and warm boot, or does windows have to load before warm boot?


              Originally posted by Megatonks
              I just ran a memtest 86+ for 7 hours overnight with 6 passes (6 and a half) and no errors... that was with these 3 sticks in. All find and dandy.

              Will memtest detect RAM errors if they're not directly caused by RAM faults? eg... if BIOS settings aren't suitable and cause errors etc?
              Sometimes it might.
              I haven't used Memtest86+ for quite long time, does it have multi CPUs option? Memtest (without the +) does, it may be more like what the OS.


              Originally posted by Megatonks
              (won't embed, too large)
              You can run multiple CPU-Z instances, put the tight next to each other and HWMonitor, launch Snipping Tool, select the area you want and snip, save, open with Paint clear background if any, save, and there you have a much smaller image.


              Originally posted by Megatonks
              Would damaged pins in the slots be something i'd be able to see do you think? although the old RAM has been fine! hmmm
              Would damaged pins pass Memtest86+ or will they be fixed with warm boot?
              If it has anything to do with the pins, human eye won't be able to see it.


              Originally posted by Megatonks
              Hypothetically, what will I need to change when adding the other 3 sticks to try 12 gb again? Just DRAM Voltage (maybe) and the QPI? I took the QPI fairly high last time and it didn't achieve stability...but maybe there was something else going on.
              QPI likely and maybe DRAM Voltage.

              Comment


              • #8
                Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                Originally posted by Chike View Post
                Yeah that was the idea, just to make sure this still works.
                Suppose you first boot until post/BIOS/boot menu, and then reboot would there be errors? I mean how long should it be between cold and warm boot, or does windows have to load before warm boot?

                Sometimes it might.
                I haven't used Memtest86+ for quite long time, does it have multi CPUs option? Memtest (without the +) does, it may be more like what the OS.

                You can run multiple CPU-Z instances, put the tight next to each other and HWMonitor, launch Snipping Tool, select the area you want and snip, save, open with Paint clear background if any, save, and there you have a much smaller image.

                Would damaged pins pass Memtest86+ or will they be fixed with warm boot?
                If it has anything to do with the pins, human eye won't be able to see it.

                QPI likely and maybe DRAM Voltage.
                Usually if it's a cold boot it can get to windows log-in screen. If I immediately go ahead and try to use it, it'll BSOD (sometimes it BSOD anyway, but then i repeat and try again until it sits patiently) so I don't touch it for 30mins or so until it's 'warmed up' so to speak, then it doesn't BSOD on me. Very bizarre. Maybe it's a coincidence or placebo, but it does seem to make a difference!

                My rig posts and gets to windows loading absolutely fine every time, just at windows loading stages i'm getting issues.

                Thanks for CPU-Z tips, didn't know you could launch multiple instances!

                Memtest86+, from what i've read, is the best one to use, so i imagine it covers all bases? The configuration menu in memtest is mostly gobbledygook for me so I just let it do it's thing.

                I'm currently using F7L BIOS for my mobo. http://www.gigabyte.com/products/pro...?pid=3305#bios.

                Should I try the newer one (beta?) or roll back to the previous non-beta? Seems strange it's such an old beta without being confirmed as stable.

                I'll try adding the extra sticks one-by-one tonight to see at what point it gives up. I'll slowly raise QPI to try encouraging it!

                Comment


                • #9
                  Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                  So maybe it is physical contact issue, maybe 0.01 mm in module thickness makes a difference.
                  If you turn the PC off after being on for a while and then cold boot right away will it behave like warm boot?
                  Try four modules in slots 1,2 3, 4. Slot 5 is the only one we can try to eliminate unless 2, 4, 6 configuration is possible.
                  Doesn't look like BIOS issue the way you describe it, I would test old RAMS first.
                  Does Memtest give error when cold booted, and if it does, what addresses?

                  Comment


                  • #10
                    Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                    Originally posted by Chike View Post
                    So maybe it is physical contact issue, maybe 0.01 mm in module thickness makes a difference.
                    If you turn the PC off after being on for a while and then cold boot right away will it behave like warm boot?
                    Try four modules in slots 1,2 3, 4. Slot 5 is the only one we can try to eliminate unless 2, 4, 6 configuration is possible.
                    Doesn't look like BIOS issue the way you describe it, I would test old RAMS first.
                    Does Memtest give error when cold booted, and if it does, what addresses?
                    Perhaps! Sigh.

                    Yes, I can shutdown my pc after being on for a while, then it'll boot back on A-OK with no issues. It's as if it physically being warm helps somehow haha?!

                    Hmmm... memtest on cold boot.... I'll do that next time I do some tests! Make's sense I guess. Although up until now I've figured 'If the ram's faulty, it's faulty..."

                    Just for the record, when I refer to 'cold boot', and I could be using the term wrong here, I mean when it's been off for a while and physically at room temperature again (most likely).

                    I think my motherboard's manual is fairly strict with which slots to use with different amounts of sticks.

                    If I were to use 5 sticks, which slots would I use? 1,2,3,4,5? Would it then only run in single-channel? To be fair, i'm not that bothered by single/dual/triple... apparently there's such a small difference in 95% of uses.

                    Comment


                    • #11
                      Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                      Originally posted by Megatonks View Post
                      Yes, I can shutdown my pc after being on for a while, then it'll boot back on A-OK with no issues. It's as if it physically being warm helps somehow haha?!
                      From the old days I went to school and actually got in class sometimes I remember that objects expand when their temperature rise. It is very possible that a not 100% tight connection id fixed when things heat up.

                      Originally posted by Megatonks
                      Hmmm... memtest on cold boot.... I'll do that next time I do some tests! Make's sense I guess. Although up until now I've figured 'If the ram's faulty, it's faulty..."
                      That is true, if... it's the RAM that is faulty.


                      Originally posted by Megatonks
                      Just for the record, when I refer to 'cold boot', and I could be using the term wrong here, I mean when it's been off for a while and physically at room temperature again (most likely).
                      Cold boot = by power from shutdown, warm boot = reboot, but we need to find if it's the "cold boot" or actually the cold.


                      Originally posted by Megatonks
                      I think my motherboard's manual is fairly strict with which slots to use with different amounts of sticks.

                      If I were to use 5 sticks, which slots would I use? 1,2,3,4,5? Would it then only run in single-channel? To be fair, i'm not that bothered by single/dual/triple... apparently there's such a small difference in 95% of uses.
                      They all are I think, ever since 2nd gen.
                      There is no 5 sticks configuration in the book but there is 4(dual channel).
                      Can you post/attach both new and old RAM SPD data? You can get this with Thaiphoon, maybe with some luckm if old RAM work in slots 1, 3, 5, you can get them to work together for 12GB total.

                      Comment


                      • #12
                        Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                        Originally posted by Chike View Post
                        Can you post/attach both new and old RAM SPD data? You can get this with Thaiphoon, maybe with some luckm if old RAM work in slots 1, 3, 5, you can get them to work together for 12GB total.
                        I've installed Thaiphoon and here's the 3 snapshots for the NEW mushkin ram (not my old 100% stable corsair xms3)


                        at 50h



                        at 52h




                        at 54h



                        I assumed there are 3 readings because there's 3 sticks, but i'm not certain, so apologies if I've posted 3 for no reason!

                        I'm gonna leave my pc off overnight then run memtest first thing in the morning all day on these current 3 sticks.

                        If I know they're fine, then i'll try with 6 sticks the following day!
                        Attached Files

                        Comment


                        • #13
                          Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                          Yes, only one of them they are all the same, and one of the others for comparison.

                          Comment


                          • #14
                            Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                            AAaaAAArrghghgrhghghgrghr

                            This is driving me insane!

                            Just spent the past few hours trialling and erroring memtests with diff BIOS settings and swapping around DIMMS and slots and what-not.

                            At first I had the 6gb (3 dimms) stable. Yay.

                            Then added one more and moved into dual channel...passed a 30minute memtest, 8gb, yay (good enough for now)

                            Added a 5th stick into the 3rd triple channel slot. All picked up by bios fine, 10gb recognised, but was giving memtest errors immediately: Addresses always between 7-8GB memory

                            Did some Bios tweaking, relaxing timings to 11-11-11-24 and slightly under 1066mhz, increasing voltages and tinkering with the uncore (common 'successful changes' were between x13 to x15... my SPD is at 6.) ..DRAM to 1.66v and QPI to 1.42v maximum at one stage... and got these 5 sticks down to only one error in 10 tests, rather than the errors on each test (all in the same place & address range) i was getting before.

                            Figured this was hopeful, and good progress, so went for the 6th stick. Errors started happening again. Nothing I could change in Bios seemed to help, and I didn't want to up the QPI anymore to be honest.

                            Took the 2 out to leave me with just 4 sticks in dual channel... fine with no errors (no immediate ones anyway...like before)

                            So I figure... the errors are in the same range every time, the 7-8GB range tells me it's either the DIMM in slot 4, or an error with slot 4 itself? Some research indicated this may well NOT be the case as it doesn't always sequence that way, so I did some swapping and trialling to try and pinpoint the Faulty DIMM.

                            However...None of them seemed to be the specific cause....strange..Maybe it's a slot?

                            Slot 1 and 2. fine, no errors.

                            Slot 1, 2, 3, 5 and 6.... no errors. 10 gb fine... So the issue must be slot 4 right? Yay, pinpointed the problem (although still annoying)...

                            Lets pop the last stick into slot 4 to confirm... run memtest...wait.. NO ERRORS

                            WHAT?!

                            I'm now booted into windows with all 12gb that randomly stopped giving the immediate errors that was happening during the initial memtests. Great...if it lasts. If the last few weeks for me are anything to go by, cold boots don't work well for me!

                            The only thing I can think of is that my multiple installing and removal of DIMMS into slots i'd never used before (2,4 and 6) has 'cleared' them or something?

                            Or Maybe it's just that with all the testing it's toasty and warm again and tomorrow morning i'll not be able to boot again. I'll report back with regard to that.

                            SO i'm now booted with 12gb at

                            QPI of 1.4v (I think)
                            DRAM at 1.66v
                            XMP off (no point when i'm setting my own DRAM and QPI voltages right?)
                            SPD at 6
                            BCLK at 179, 20 Multi for my overclock (In the middle of my testing I figured why not, and it didn't seem to make a difference), CPU at 1.3v. Turbo OFF
                            That BCLK is putting the memory at JUST under 1066, as 1333 was giving hundred of errors in seconds.
                            11-11-11-24 timings at 2T command rate.

                            None of this has followed any patterns at all for it to make sense or to pinpoint anything really. Unless you guys can see something I cant?

                            Here's some screenies of CPU-Z and Thaiphoon and HW monitor again

                            Comment


                            • #15
                              Re: X58A-UD3R rev 1.0 &amp; Xeon X5650 &amp; Mushkin 12gb pc3-12800 - UNSTABLE

                              1066 CAS 11? I can move data faster by hand even if I need to do it bit by bit.
                              When did 1333 gave errors, would it error if you set it to 1333 after the PC has been on for a while?

                              Comment

                              Working...
                              X