Announcement

Collapse

Password resets

See more
See less

BSOD Mystery - i5 6600, G1 Gaming 7, Vengeance LPX 2666

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

  • BSOD Mystery - i5 6600, G1 Gaming 7, Vengeance LPX 2666

    System:
    i5 6600 (non k) Skylake
    Gigabyte G1 Gaming 7 Z170X
    Vengeance LPX 2x8GB 2666 (mobo downs it to 2133)
    SanDisk 128GB SSD
    Corsair CX750 Modular PSU
    Sapphire R9 280X Vapour-X
    Windows 10, Ubuntu 16.04

    Around March last year I decided to upgrade my PC with a brand new mobo, CPU and RAM moving to Skylake and DDR4, I noticed a nice email from aria with a deal for the G1 Gaming 7 mobo with a 750w modular PSU so I went for it.

    The new system was great, until November where the first BSOD happened, IRL_NOT_LESS_OR_EQAL, I put this down to unoptimized ARK and just went on and all was fine until a few weeks later when the BSODs slowly became progressively more frequent. ATTEMPTED_EXECUTE_OF_NOEXECUTE_MEMORY, MEMORY_MANAGEMENT, PAGE_FAULT_IN_NON_PAGED_AREA were the most common. Eventually I was unable to boot completely.

    So I began the process of elimination, I tried each stick of RAM individually in each slot as well as removing the GPU to no avail. I then decided to take the mobo out and put it into my old PC with a fresh Win 10 installation, a different PSU, a 2TB Barracuda HDD and no GPU but still the same problems.

    I decided that with CPUs rarely going bad and the odds of both sticks being bad unlikely(?) that my best bet is to RMA the mobo. I sent it away only to receive it back with the CMOS reset and updated and no faults found. So I put everything back together and gave it another try only to meet the same BSODs as before.

    So I ran memtest overnight, no problems found. This is when I decided to boot into Ubuntu to see how it would react. On booting it was stuck on a console spamming something about the cpu/core rate being so many ms off, I'll rerun this test and will update this post with the exact message.

    So I decided to play around in the Frequency Settings. The first thing I did was disable Intel Turbo Boost, after this I successfully booted with no issues at all! But, my CPU was reduced down to a measly 0.8ghz! I tried to override the clock speed from the auto 8x multiplier up to the default 33x but then my PC wouldn't boot at all so I had to reset the CMOS. Next time I kept Intel Turbo Boost enabled but lowered it to 33x on each core, kind of a soft disable I guess and this worked great.

    But now, just into 2017, they're back and getting progressively worse again! I've lowered the multiplier down to 31x and things seem to be stable for now. I don't have a clue what this could be, I've tried enabling XMP and that didn't change anything either. Temps are all good too. I also ran that Intel Diagnostic Tool and it came up with no errors, though I was running it under my 31x multiplier as I can't even boot at 32+. By default Turbo Boost has core 1 at 39 down to core 4 at 36.

    I have one theory about the RAM being at 2666 and maybe the mobo isn't downing it to 2133 correctly or something, but to test this I'll need to buy more RAM as the current RAM seems to be good as far as memcheck is concerned but DDR4 RAM isn't cheap so this is a risk. If not, the CPU is bad?

    The random BSODs (that started initially and then later with turbo nerfed to 33x) happened when idle or when gaming, though there is a pattern as the last time this happened was when I was running an ARK Server and playing it at the same time (a lot of RAM usage) and now I'm running an Empyrion Server and playing it at the same time, but again, this also occurs when idle and sometimes within a few minutes after a clean boot.


    TL;DR BSODs in Win 10, tried fresh install, different PSU, Drives, individual RAM sticks (don't have spare DDR4 RAM), removed GPU. Memcheck overnight all good. Works with Intel Turbo off but then stuck at 0.8ghz, also works with Turbo on but limited to 31x. Intel Diagnostics good (under lower clock). Ubuntu spits out core speed errors. Temps good, BIOS up to date, mobo returned from RMA with no faults. Maybe RAM not lowering from 2666 correctly or bad CPU?

    I'll get the exact Ubuntu output and will add it to this post.

    BSODs:
    Bugcheck code: 0x3B (0xC0000005, 0xFFFFF802F7441D3B, 0xFFFFD000235F2C80, 0x0)
    Error: SYSTEM_SERVICE_EXCEPTION
    file path: C:\Windows\system32\ntoskrnl.exe

    Bugcheck code: 0xA (0xFFFFD00025BD9D60, 0x2, 0x1, 0xFFFFF80102455778)
    Error: IRQL_NOT_LESS_OR_EQUAL
    file path: C:\Windows\system32\ntoskrnl.exe

    Bugcheck code: 0xA (0x8, 0x2, 0x0, 0xFFFFF801D1806CB6)
    Error: IRQL_NOT_LESS_OR_EQUAL
    file path: C:\Windows\system32\hal.dll

    Bugcheck code: 0x00000050ffffd000`20c5a720, 00000000`00000002, fffff800`e97d5f4f, 00000000`00000000
    Error: PAGE_FAULT_IN_NONPAGED_AREA
    file path: C:\Windows\system32\ntoskrnl.exe
    Last edited by Drahkir; 16-01-17, 17:05.

  • #2
    2133 is the default, failsafe speed for ddr4. all ram will run at that speed till you change the speed/latency/voltages manually or with xmp.

    i would start by pulling the cmos bat, draining the power and leaving it overnight. then with just the cpu and ram in and using the igp i would try to load the bios. set it to optimised defaults, save, restart, back in bios to enable xmp profile for ram, check boot order and set other devices as needed for your system.

    assuming that boots and sits in the bios ok without lockups, i would then look at loading an os. windows or linux both have testing and monitoring software so which matters not, but finding out the system temps and running voltages does.

    have you tried to scan the drive and fix any errors it has?

    could be nothing to do with the back bone and might just be a dodgy drive?

    to me it looks like you have narrowed it down to a cpu/ram issue, but it could just be a config issue too. only testing and taking notes of what it does will show you, without dupe parts for everything to swap around.
    "Those able to see beyond the shadows and lies of their culture will never be understood, let alone believed, by the masses."
    Plato

    Comment


    • #3
      Originally posted by luke22 View Post
      2133 is the default, failsafe speed for ddr4. all ram will run at that speed till you change the speed/latency/voltages manually or with xmp.

      i would start by pulling the cmos bat, draining the power and leaving it overnight. then with just the cpu and ram in and using the igp i would try to load the bios. set it to optimised defaults, save, restart, back in bios to enable xmp profile for ram, check boot order and set other devices as needed for your system.

      assuming that boots and sits in the bios ok without lockups, i would then look at loading an os. windows or linux both have testing and monitoring software so which matters not, but finding out the system temps and running voltages does.

      have you tried to scan the drive and fix any errors it has?

      could be nothing to do with the back bone and might just be a dodgy drive?

      to me it looks like you have narrowed it down to a cpu/ram issue, but it could just be a config issue too. only testing and taking notes of what it does will show you, without dupe parts for everything to swap around.
      Thanks for the response! I'll try pulling the battery and see what happens though when I received the mobo back from RMA Gigabyte had updated and cleared it all, but I'll try anyway to make sure. The strange thing is the problem slowly started out of nowhere and got progressively worse without me touching any settings.

      I've ruled out dodgy drives as I've had the mobo, RAM and CPU all in my older tower which has a different CPU and a spare drive, I was also running Ubuntu form a Live USB.

      For temps my CPU idles around 20-25C with the occasional spike to near 30C, when gaming it creeps to 50C. My GPU sits at around 40C but can reach around 70C when gaming, however I can rule this out as I have issues without it and it works fine in my old system. I'll monitor the voltages as that is something I haven't done yet.

      The system has never locked up at BIOS, it was also able to run Memcheck fine despite being unable to boot into Win10 (current install on SSD and clean install on HDD) or Ubuntu (Live USB).

      I'm on the system, gaming, video editing and encoding right now without any problems but this is with Turbo Boost capped at a 31 multiplier, so maybe it could be voltages I'm not too experienced with this though but I'll see what I can find without borking anything.
      Last edited by Drahkir; 18-01-17, 13:28.

      Comment


      • #4
        Trying with a reset CMOS (not left overnight yet) and I've come across two new BSODs, they don't seem to leave a dmp either:

        KERNEL_AUTO_BOOST_INVALID_LOCK_RELEASE

        MEMORY1_INITIALIZATION_FAILED

        Again, capping Turbo Boost to 31x makes the system stable again (at least so far). Now my CPU is a locked non-k 6600, but it should be able to Turbo Boost up to 3.9ghz with a base of 3.3ghz, yet everything seems to be unstable unless I cap at 3.1ghz, 3.2ghz seemed ok at first but didn't last long.
        Last edited by Drahkir; 18-01-17, 13:28.

        Comment


        • #5
          Weird.

          Might be your CPU is bad. It's possible. My 2600k used to be clocked at 4.8ghz and was fine for ages, then slowly started getting bsod's. I dropped it down a bit and it was fine. Temps were always ok, so I guess it's possible for a CPU just to degrade.

          It sounds like yours may be a more serious case if you have to significantly down clock even lower than stock base speeds. Looking at your posts it appears to be the one variable that stops the bsod's.
          GIVE PEAS A CHANCE

          Comment


          • #6
            Originally posted by BUDFORCE View Post
            Weird.

            Might be your CPU is bad. It's possible. My 2600k used to be clocked at 4.8ghz and was fine for ages, then slowly started getting bsod's. I dropped it down a bit and it was fine. Temps were always ok, so I guess it's possible for a CPU just to degrade.

            It sounds like yours may be a more serious case if you have to significantly down clock even lower than stock base speeds. Looking at your posts it appears to be the one variable that stops the bsod's.
            Yeah I'm thinking it could just be the CPU, it's not even a year old yet but I could just be really unlucky. One more thing I want to try first is setting everything to my cpu stock settings rather than leaving them on auto, just to rule out my mobo getting the auto rates wrong or something.

            Comment


            • #7
              Well I got my BSOD again. Instead of dropping to 3.0ghz I decided to restore everything back to defaults, again BSODs as usual so I decided to try keeping at the default 3.3-3.9ghz but reduce the cores 3 and everything seems to be ok again.
              Obviously running with only 3/4 cores is no good at all, it looks like I might have to RMA the CPU. If the system is usable and performs well I'll wait it out until near the end of March as I hope to try out the new AMD Ryzen stuff, once I get that I can RMA without having no PC.

              I did RMA the mobo and that came back find but I'm still sceptical as I've heard over at the pcmasterrace reddit that you can't trust mobo RMAs, but there's nothing else I can really try so I'll just bite the bullet and RMA the CPU when I can.

              Comment


              • #8
                have you tried speaking to gigabyte?

                while i would not rule out the cpu, i would check for any known issues with your system parts or maybe a bios update?
                "Those able to see beyond the shadows and lies of their culture will never be understood, let alone believed, by the masses."
                Plato

                Comment


                • #9
                  That's a good idea, I'll try contacting them about it. It looks like it's probably some form of hardware failure though as the situation has slowly been getting worse over time, but still worth a try. I'm going to be getting a new Ryzen CPU and a mobo to go with it soon, so as long as my system lasts out until then I can RMA away.

                  Comment


                  • #10
                    I got my money on the CPU, let us know though. These kind of threads are good for building up knowledge.
                    GIVE PEAS A CHANCE

                    Comment


                    • #11
                      Just an update, I still haven't got round to contacting Gigabyte, I'm not using my PC for anything major performance wise currently so I've kind of stopped caring. The Ryzen 5 CPUs are coming out in April so I'm going to start looking at getting a new mobo in preparation for that now that I have some cash saved up.

                      Once my new Ryzen system is up and running I'll then send the 6600 away for RMA to see what happens and will report back on the outcome. If it is the problem and is replaced I'm going to donate it to my brother as his current system is really dated, if it's returned with no faults though then I'll be at a loss, I might just sell off the CPU and mobo seperately.

                      Edit: My system is still running fine but at 3 cores.

                      Comment


                      • #12
                        An update! I have a brand new system now with a Ryzen 5 1600X cpu and Gigabyte AX370 Gaming K 7 mobo with a fresh install on a new SSD. The gpu (R9 280X) is the same and the RAM is the same.

                        All seemed well until I started test recording ARK and Empyrion when all of a sudden, the same random BSODs have returned!

                        The only consistent factor is the RAM, so I've ordered more RAM which should arrive by Thursday at the latest, if the new RAM fixes my old system and my new system then I can confirm it is the RAM all along. This would also mean that memtest cannot be trusted.

                        Another worry is that my PSU could be slowly damaging hardware so that even when I test with my old PSU the same problems occur but because the damage is already done, is this thing, can a faulty PSU slowly damage hardware over time like this?

                        If the same problems occur with my new RAM then I'm at a loss and have two systems with the same problem!
                        Last edited by Drahkir; 18-04-17, 10:13.

                        Comment


                        • #13
                          I've got new RAM and sure enough, the problem has gone away. So despite the RAM passing MemCheck it was still the problem. The strange thing is that BOTH sticks were faulty and they were gradually getting worse. On testing individual sticks I found that one stick lasted a lot longer before blue screening than the other with the Ryzen system, in my Skylake system there is a BSOD on boot with both sticks.

                          My next concern is how the RAM degraded, could both sticks just have a flaw? I just hope that it wasn't something like the PSU slowly damaging the RAM as I'm using this with my new RAM in my Ryzen system now or could the mobo have done this, though it was returned from RMA with no faults.

                          Anyways I'm going to return the faulty RAM giving me up to 32GB so at least there's that. My old Skylake CPU and Mobo are being donated to my brother who's in dire need of an upgrade.

                          Hopefully this is the end of my several month long BSOD ordeal! If not then the PSU must be damaging the RAM in which case I can RMA the RAM again and replace the PSU with my spare, only time will tell!

                          Comment


                          • #14
                            Interesting indeed, and that memtest did not pick anything up, also looks like i was wrong with my CPU theory.
                            GIVE PEAS A CHANCE

                            Comment


                            • #15
                              glad you have found the issue dude

                              maybe it is time for a new ramtest?

                              https://www.techpowerup.com/memtest64/
                              "Those able to see beyond the shadows and lies of their culture will never be understood, let alone believed, by the masses."
                              Plato

                              Comment

                              Working...
                              X