Warning: Some i7-6820HKs and i7-6700HQ have Uneven Core Temps due to Uneven Heatsink

Discussion in 'Hardware Components and Aftermarket Upgrades' started by iunlock, Oct 25, 2016.

?

Does your 6820HK have cores that are 5c or higher than the others?

  1. Yes, I have a 6820HK that has 1 or more cores with a high temperature differential, greater than 5c.

    14 vote(s)
    9.5%
  2. Yes, I have a 6820HK that has 1 or more cores with a high temperature differential, greater than 10c

    21 vote(s)
    14.2%
  3. Yes, I have a 6820HK that has 1 or more cores with a high temperature differential, greater than 15c

    9 vote(s)
    6.1%
  4. Yes, I have a 6820HK that has 1 or more cores with a high temperature differential, greater than 20c

    23 vote(s)
    15.5%
  5. My 6820HK core temperatures are all tracking the same, with less than 5c difference.

    18 vote(s)
    12.2%
  6. Yes, I have a 6820HK where Core #3 (4th Core) is 10C cooler than the rest.

    1 vote(s)
    0.7%
  7. Yes, I have a 6700HQ that has 1 or more cores with a high temp differential, greater than 5c.

    13 vote(s)
    8.8%
  8. Yes, I have a 6700HQ that has 1 or more cores with a high temp differential, greater than 10c.

    15 vote(s)
    10.1%
  9. Yes, I have a 6700HQ that has 1 or more cores with a high temp differential, greater than 15c.

    10 vote(s)
    6.8%
  10. Yes, I have a 6700HQ that has 1 or more cores with a high temp differential, greater than 20c.

    10 vote(s)
    6.8%
  11. My 6700HQ core temperatures are all tracking the same, with less than 5c difference.

    12 vote(s)
    8.1%
  12. Yes, I have a 6700HQ where Core #3 (4th Core) is 10C cooler than the rest.

    3 vote(s)
    2.0%
  1. Falkentyne

    Falkentyne Notebook Deity

    Reputations:
    1,079
    Messages:
    1,479
    Likes Received:
    1,462
    Trophy Points:
    181
    The GT73 has fully separate heat pipes, but the GPU VRM heatsink is cooled by the CPU fan, but the CPU and GPU heatpipes themselves are fully separate and do not connect. You can repaste the CPU or GPU separately without having to remove the other heatsink (you do have to remove both fans to repaste the GPU, due to the VRM heatsink being cooled by the CPU fan, but not to repaste the CPU).

    The GT72 has a heatpipe on the CPU go to the GPU fan (which is rather strange), and I'm not sure if the heatsink that pipe goes to is shared with the GPU heatsink or not (if it is, you would have to remove both heatsinks to repaste...a HUGE pain in the butt).
     
    hmscott likes this.
  2. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    3,631
    Messages:
    13,543
    Likes Received:
    15,682
    Trophy Points:
    931
    You're blind then :)

    Not only myself, but hundreds of thousands or more other people have no problems out of the box.

    We see all the real problems here, of the people that are aware of such resources, and many of those are OCD'ing to the max due to being hypersensitive to such issues.

    Many go away happy with no physical intervention necessary, they only needed undervolting and proper interpretation of data - looking at all the points of monitoring between temperature peaks.

    Most people don't need to delid their processors, unwarp their heatplates, and wait months for BIOS/EC hacks from Prema for their laptops - you on the other hand live 100% in that world.

    So of course you would think everyone has problems with their laptops out of the box, because in the Clevo world, they do :)
     
    Last edited: Jul 20, 2017
    Aman Krishna likes this.
  3. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    3,631
    Messages:
    13,543
    Likes Received:
    15,682
    Trophy Points:
    931
    The difference between persistent - which is what I have been encouraging you do be - and OCD is that with persistence there's a good reason for it.

    With OCD it's often completely imaginary, or overblown. We get people like that in here all the time. Their temps are fine, but in their mind any reduction in temperature they can accomplish is an imperative that must be followed to its iterative end.

    It's a fine line to draw, but obvious when you see it :)
     
  4. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    3,631
    Messages:
    13,543
    Likes Received:
    15,682
    Trophy Points:
    931
    Yes!!

    This is showing improved evening pressure of the heatplate against the CPU!

    You're temperature differential is about 1/4 what it was to start!!

    You can stop now if you want. :)

    At least take a break and think about what you saw and what you did, something might jump out in your mind that might be still keeping the heatplate from applying 100% even pressure.

    As with all problems, the "last" fix doesn't always happen right away ;)
     
    Last edited: Jul 20, 2017
    Aman Krishna likes this.
  5. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    3,631
    Messages:
    13,543
    Likes Received:
    15,682
    Trophy Points:
    931
    That's great, low enough to leave as is, 2c-3c differential :)

    Try Prime95 small fft and see how much further the differential jumps up...
     
    Last edited: Jul 21, 2017
    Aman Krishna likes this.
  6. Falkentyne

    Falkentyne Notebook Deity

    Reputations:
    1,079
    Messages:
    1,479
    Likes Received:
    1,462
    Trophy Points:
    181
    Heh I'm too busy trying to flash my video card with an SPI flasher to try to trick the GT73VR from changing from a 7RE to a 7RF....(you can see my post in the pascal bios editor thread).
    I was able to run OCCT between disassembles....
     
    hmscott likes this.
  7. plee82

    plee82 Notebook Evangelist

    Reputations:
    90
    Messages:
    488
    Likes Received:
    338
    Trophy Points:
    76
    Even at 5C I will stop. I feel like if I replace every single pad with k5 pro, the HS will even out 100%. However, I am not sure I wanna go that direction lol.
     
    hmscott likes this.
  8. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    3,631
    Messages:
    13,543
    Likes Received:
    15,682
    Trophy Points:
    931
    Congratulations on your successful problem resolution, and OCD avoidance :)

    I'm pleased this resolved so quickly, with such a problem it can take a longer time to find them. This thread started out that way, and took a while before the thermal pad height was noticed as being "a thing".

    That's why threads like this are so important, to pick up and set new people on the right path quickly.

    Maybe you could make a thread in your Brand forum, or a post in the Owners thread for your laptop with the problem and resolution, you've already got the info fresh in your head, and photo's of the problem too. I see you've already started, that's great.

    Thanks for hanging in there :)
     
    Last edited: Jul 20, 2017
    tilleroftheearth and Papusan like this.
  9. plee82

    plee82 Notebook Evangelist

    Reputations:
    90
    Messages:
    488
    Likes Received:
    338
    Trophy Points:
    76
    Yup I have been communicating my progress to the GT62VR thread since @encheels has been having temp diff issues. Thanks for the K5 Pro recommendation. It works really well. By the way, what are symptoms of VRM overheating? Does the CPU throttle? Just curious.
     
    hmscott likes this.
  10. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    3,631
    Messages:
    13,543
    Likes Received:
    15,682
    Trophy Points:
    931
    I haven't seen laptop CPU VRM problems, but I have seen GPU VRM problems, usually because the VRM's aren't touching the heatplate - have no paste or thermal pad.

    I wouldn't worry about this if you aren't having any performance issues with the GPU.

    While you had the GPU heatplate off, did you check the VRM's for coverage issues? Check your photo's, maybe you got lucky and captured the coverage with your snapshots.

    I don't think it's worth pulling things apart again to check this, unless you develop GPU flakiness issues.

    The Intel x299 CPU VRM issues haven't even been seen in motherboards for a long time - or very often, so I wouldn't map those onto your laptop.

    Intel's x299 i9 spec lapse due to last minute extension of the i9 CPU line is behind the x299 issues, along with not enough testing time available for the x299 motherboard manufacturers due to Intel's rush to market.

    For now I think you can stop worrying about issues like this :)
     
Loading...

Share This Page