Official Intel Haswell (K) Overclocking thread

Discussion in 'Processors and motherboards Intel' started by -Tj-, Jun 7, 2013.

  1. dove

    dove Master Guru

    Messages:
    312
    Likes Received:
    0
    GPU:
    Powercolor 6950 2GB
    Interesting. I don't use RAID and I can't see any reason to suspect my RAM is a bit wonky (I'd expect crashes/BSOD more often if that was the case, not just when transferring data between two USB3 drives) but it might be worth a try uninstalling and reinstalling all the drivers. Thanks for the suggestion.

    I was going to try a fresh install of Win 8.1 but I don't currently have space on my SSD for a second install (don't want to mess with the first one until I've tested to see if the fresh one makes any difference) and I can't face installing and running Windows on a HDD partition after using a SSD ;)
     
  2. dove

    dove Master Guru

    Messages:
    312
    Likes Received:
    0
    GPU:
    Powercolor 6950 2GB
    Hell no it isn't! I have to use 1.26v to get 4.4Ghz with my 4670k, although my Dad's one only needed 1.225v.
     
  3. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    :)

    I do have a compelling argument for saying it's 1.215 tho as, even tho I set the adaptive limit to 1.23 in the BIOS, all reporting software, including MSI's own Command Centre, reports the VID as 1.214 or 1.215.
     
  4. dove

    dove Master Guru

    Messages:
    312
    Likes Received:
    0
    GPU:
    Powercolor 6950 2GB
    Actually it turned out to not be that easy. The first driver I tried uninstalling was "Intel(R) USB 3.0 eXtensible Host Controller - 0100 (Microsoft)" but despite it's name, it disabled all my USB2 ports and devices, including my mouse and keyboard, so all I could do was hard reboot.

    On rebooting, it automatically reinstalled the drivers for that and all the devices I have connected to my USB 2 ports but perhaps it will be easier and safer just to image the OS partition again and do another Universal Restore, which should strip out all the drivers and re-install them fresh, rather than manually uninstalling drivers one by one, with the risk of causing issues like this.
     

  5. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    Just had a restart during CS:GO (no BSOD) which I presume is due to me bumping the uncore up to 4.2Ghz.

    I was on 1.088v on ring so I've bumped that to 1.35 to see of that helps.

    I presume I'm on the right lines here and an unstable uncore can/will cause restarts?

    If not, there's clearly something else wrong but it didn't restart when it was on 3.9Ghz.
     
  6. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,624
    Likes Received:
    2,239
    GPU:
    Zotac GTX980Ti OC
    lol no dont go that crazy, set to 1.15v should enough and yep 1.10v sounded too good to be true :nerd:

    Yes sudden reboot can be cache/vring issue, sometimes not enough cpu current for extra vrm power when OC'ing.
     
  7. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    Lol, just noticed my mistake, it's set to 1.135! (1.15 in BIOS)
     
  8. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    Any way to check what setting my OCP is running at? Currently set to auto, along with everything else in that section but has percentages available from 12.5% all the way to 200%
     
  9. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    I've just thought of something. Personally, I can't believe I've missed this until now.

    My psu is bottom mounted with its fan facing down, but... My case is on carpet.

    In the 2 months of owning this new setup I've only ever had about 2 bsod restarts and they were all known issues when I was experimenting with voltages.

    The psu overheating would pretty much explain everything, especially with the restarts only happening with games as the psu is working harder and thus getting hotter.

    I'm gonna try putting something under it tomorrow, bed time now.

    Also means I might be able to lower some voltages on this overclock too as I only raised them to avoid restarts in games. Never had one while only stress testing the cpu.
     
  10. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,624
    Likes Received:
    2,239
    GPU:
    Zotac GTX980Ti OC
    nah sudden reboots are not psu related if you know you're ok with a bit lower OC.
    Haswell doesn't consume much.. Your whole system is probably max 350-400w.


    When you OC cache with low voltage you can expect sudden reboots, wmea 0x124 bsod or sometimes just hard freeze., higher cache voltage usually fixes it.
    Well there is a wall by certain ratio, it is not meant to be in sync with cpu frequency, now its separate internal NB or QPI like by old Nehalems.
     
    Last edited: Oct 26, 2014

  11. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    I'm not so sure to be honest, does seem odd that it only crashes when running games. It's never once restarted using CPU stress testing, yes I've had 0x124 errors when I was finding a stable voltage but never a hard restart.

    More voltage == more watts == more heat, both in the PSU and on the actual component.

    I've still not entirely ruled out the gfx card is faulty but I raised the back end of my PC up a few mm and played CS:GO for a couple of hours this afternoon without a reboot. Ring voltage is at 1.135 tho.
     
    Last edited: Oct 26, 2014
  12. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,624
    Likes Received:
    2,239
    GPU:
    Zotac GTX980Ti OC
    See, 1.10v was obviously too low for 4.2ghz.

    Those sudden reboots happened by me too when I OC'ed cache/vring/uncore e.g. 4.2ghz to 4.4ghz at lower voltage,
    too low - it reboots
    somewhere half stable - it bsod wmea 0x124
    almost stable - it freezes
     
  13. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    I see where you're going on that one.

    Personally I'd rate them in order of severity thus:
    Reboot > freeze > BSOD

    Reason being is a BSOD is where the system knows something is wrong but still has enough faculties about it to generate the BSOD, which is a RING0 kernel event, the system is technically still "running" at this point, but the but the other 2 events are so catastrophic it doesn't have the chance to do this.

    If you get a freeze/reboot then increase the voltage and get a BSOD instead you're halfway there.
     
  14. Fender178

    Fender178 Ancient Guru

    Messages:
    4,198
    Likes Received:
    213
    GPU:
    GTX 1070 | GTX 1060
    Well I thought I fixed my odd problem with my computer turning off after a few seconds on the Windows 8.1 loading screen resetting my OC and settings back to defaults but it did it again so I tried resetting the rest of my settings that I tinkered with the CPU power states back to auto to see if that fixes my problem because I have a theory that every time you turn on your PC there is a tiny bit of a surge of power because of everything turning on and I feel that the CPU power states is interfering with something with the process with Windows when it first starts up power wise.
     
  15. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,624
    Likes Received:
    2,239
    GPU:
    Zotac GTX980Ti OC
    ^
    By win logon it activates your OC, if not stable it resets etc.. I think its nothing to do with C-states.


    If you want to have max OC by boot you need to select that in bios apply max turbo by boot, otherwise it always boots by max non turbo 4.0ghz or in my case 3.5ghz.
     

  16. Fender178

    Fender178 Ancient Guru

    Messages:
    4,198
    Likes Received:
    213
    GPU:
    GTX 1070 | GTX 1060
    Well something is causing it to shut down before Windows fully loads. So I am setting all of my Bios settings to stock setting as a method of troubleshooting.
     
  17. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,624
    Likes Received:
    2,239
    GPU:
    Zotac GTX980Ti OC
    imo its your OC is not stable when it activates its OC multi.


    It boots @ 4ghz, right by win logon it activates your OC e.g. 4.6ghz but its either not enough cpu voltage or cpu power current in DIGI+, I think its just cpuv issue.


    Well you can keep c-states @ auto to eliminate that for now, just make sure you use balanced plan in idle or it will run @ max freq./fixed voltage all the time.
     
  18. Fender178

    Fender178 Ancient Guru

    Messages:
    4,198
    Likes Received:
    213
    GPU:
    GTX 1070 | GTX 1060
    I do have my plan set to balanced mode. Also I disabled my Overclock as well and it is using the turbo boost OC of 4.4ghz.
     
  19. SpecChum

    SpecChum Master Guru

    Messages:
    617
    Likes Received:
    13
    GPU:
    Vega 64 - Acer XR34
    Mine doesn't oddly enough. On the balanced plan both the frequency as voltage go down, as you'd expect but even on the performance plan, whilst the frequency is locked to 4600Mhz the voltage still goes down to 0.0whatever volts.
     
  20. -Tj-

    -Tj- Ancient Guru

    Messages:
    17,624
    Likes Received:
    2,239
    GPU:
    Zotac GTX980Ti OC
    That will happen @C7 state if you dont have enabled iRST dynamic storage accelerator and in bios - by MSI its named tinylake.
     

Share This Page