NVIDIA PR questioned over Temp increase

Discussion in 'Videocards - NVIDIA GeForce Drivers Section' started by Envoy, Mar 3, 2005.

  1. THunDA

    THunDA Ancient Guru

    Messages:
    2,739
    Likes Received:
    1
    GPU:
    Powercolor x1900xtx 675/783
    I have the same info as Emanon75... 6800gt in my Infinity rig..

    Nick .. thanks for taking the time to do this.. even if it is just miss reading on the software side it would be nice for them to adjust it back of let us know whats going on..
     
  2. V_LESTAT

    V_LESTAT Maha Guru

    Messages:
    1,297
    Likes Received:
    0
    GPU:
    PNY XLR8 GTX280
    yup i 2nd that many thanks for pushing nvidia to take notice!!
    didnt i see some people reporting this last beta fixed the temp bug for some of them ?
     
  3. MaXThReAT

    MaXThReAT Guest

    I saw a post I think it was Nvnews Forums. A guy had a beta driver with a build date of 3/8 and said NV has it corrected and any drivers dated on or after 3/8 will not have the temp bug. I wish I could find it. It looked legit.
     
  4. Grendel_66

    Grendel_66 Master Guru

    Messages:
    793
    Likes Received:
    0
    GPU:
    MSI GTX 970 Gaming 4G
    If you had payed attention you would have read that a lot people did that already and confirmed the rise.. :rolleyes: I've a temperature probe in the water circuit after the GPU and I see a rise there as well.
     

  5. Emanon75

    Emanon75 Guest

    There is no actual rise in temps.. :rolleyes:

    Check the readme ffs. :p

    EDIT:
    Open Issues in Version 71.84:
    As with every released driver, version 71.84 of the Release 70 driver has open
    issues and enhancement requests associated with it. What follows is a list of
    issues that are either not fixed or not implemented in this version. Some
    problems listed may not have been thoroughly investigated and, in fact, may
    not be NVIDIA issues. Others will have workaround solutions.
    • GeForce 6 Series, SLI: Overclocking using Coolbits does not work
    when SLI mode is enabled.
    This issue will be fixed in the next driver release.
    • GeForce 6800: The Temperature Settings; in the Display Properties
    reports an approximately 20% higher temperature than with the
    previous driver version.
    This issue will be fixed in the next driver release.


    For the last freaking time..
    There is no actual temp increase.
    It's a driver bug :rolleyes:
     
  6. Emanon75

    Emanon75 Guest

    My Compunurse temp probe tells a different story.. :p
     
  7. Mr_ALLroy

    Mr_ALLroy Ancient Guru

    Messages:
    1,994
    Likes Received:
    19
    GPU:
    EVGA RTX 3080 FTW3
    • GeForce 6800: The Temperature Settings; in the Display Properties
    reports an approximately 20% higher temperature than with the
    previous driver version.
    This issue will be fixed in the next driver release.

    That doesn't prove the temp increases are real or not. It simply states that the driver reports higher temps.

    Why then, does my cpu and mobo ambients go up only after install a 7x.xx driver? For sure there is a real temperature increase. Maybe a superficial increase for some, but for most, the temps are actually rising and causing other monitors (not GPU related) to report higher temps as well.

    And yes, I grabbed a Fluke Temp Monitor from work and confirmed, the temp increase is real.
     
    Last edited: Mar 11, 2005
  8. V_LESTAT

    V_LESTAT Maha Guru

    Messages:
    1,297
    Likes Received:
    0
    GPU:
    PNY XLR8 GTX280
    Emanon explain your exact set up and exactly how and when your seeing the increase,
    we dont care about water temps that will prove nothing. so if all your doing is going by water temps then... sorry...
    whats your actual temps using a probe on the video card.
    water temps will rise with the ambient temps of your room. Also if you have your video card on the same loop as your cpu then again water temps will rise.
    water temps are never consistant. they fluctuate too much especially if its running thru a radiator thats doing its job properly.
    The only water temps that mean anything, well ya know what,, water temps overall dont mean anything.
    the very second your water leaves the gpu block its already cooling and dropping temps.

    no offense but you judging temps raising by your water is not a smart way to judge temps.
     
  9. malorg

    malorg Member

    Messages:
    13
    Likes Received:
    0
    GPU:
    XFX 8800GT 670M XXX Edition
    hello people
    I can't see the problem going from 50 C to 60 C idle, after all, this cards starts giving messages when it reaches 120 C , don't you think you are overreacing at 60 C then ? Mine has always been 60 C idle, with 350/1000 clock or 440/1200 clock;)
    Highest temp on full load over 5-6 hours is 75 C and that's with a room temp on 24-25 C. This is NOT a major problem, it's just you people being obsessed with temps. and 60 C idle with 440/1200 on a GT card.. you tell me if that's bad. I don't think so.
     
  10. Emanon75

    Emanon75 Guest

    Well, duh.. :bonk:
    It says:
    The driver reports higher temps..

    Not that the driver increases temps..
    What part of that don't ya get? :rolleyes:

    If you say so.. :p

    I assume then, you can explain to me how overclocking, then resetting to default returns temps to normal?
     

  11. OneWord

    OneWord Maha Guru

    Messages:
    1,124
    Likes Received:
    0
    GPU:
    Gainward 7800GS BLISS 512
    The max threshold of 120.C is a joke. If your card ever reaches 120.C it will have a very short life and would die very quicky. Yea ok 60.C isnt the end of the world but if cooler is better when using air cooling.
     
  12. malorg

    malorg Member

    Messages:
    13
    Likes Received:
    0
    GPU:
    XFX 8800GT 670M XXX Edition
    heh, GPU's can handle 80-90 without a problem. They can handle a lot of heat, and nVidia set it to be 120 themself for a reason. I have the Gainward 6800 GT 2400/Ultra Golden Sample. Tho, I've never had ANY different temps with 7x drivers or the 6x drivers. Go figure ^^
     
  13. Mr_ALLroy

    Mr_ALLroy Ancient Guru

    Messages:
    1,994
    Likes Received:
    19
    GPU:
    EVGA RTX 3080 FTW3


    Why do you have such a hard time accepting a different opinion other than your own? If your not affected than good for you. For us people that ARE getting actual increases, I think you'd be better not to respond.

    And I don't make Nvidia drivers for a living so I wouldn't begin to pretend I could explain. Could you enlighten us all with your profound wisdom?
     
    Last edited: Mar 11, 2005
  14. Grendel_66

    Grendel_66 Master Guru

    Messages:
    793
    Likes Received:
    0
    GPU:
    MSI GTX 970 Gaming 4G
    You don't know my setup and you obviousley don't have any experience w/ water cooling.

    The room is airconditioned to 19C +/- .5C. I get very consistent temperature readouts and it's very reproduceable in both ways. I's just providing another datapoint for everyones convienence -- if you won't believe it, fine, your problem. Go ahead and ignore the temperature from your card :p

    Edit: the main point here was: if the driver just reports 20% higher temps, why do I see only a 5% rise ? Magic or what ?
     
    Last edited: Mar 11, 2005
  15. Emanon75

    Emanon75 Guest

    Simple enough..
    CP temps w/ 61.77 & 67.66: Average 58 - 61c @ idle.
    CP temps w/ 71.24 and up: Average 68 - 71c idle.

    I agree..
    Actual die temps need to be taken to be anywhere near accurate.

    Well, it was quite unscientific..
    Due to the heatsink on my card, I couldn't mount the probe properly, and I'm not feeling like taking the stock cooling off till my Zalman gets here.. ;)

    So.. I just shoved it up under the sink, as close to the gpu as possible..
    It read much lower than the CP, at around 45 - 48c or so, regardless of which drivers used and CP readings which is at least partly due to half assed mounting job. :p

    If there was actually a 10c increase, I'd assume that even poorly placed, it would've picked up at least a small variation, fluctuation, or increase.. However, it did not.

    So, I'm convinced it's a driver issue, at least till someone proves otherwise. ;)
     

  16. pumpkinlad

    pumpkinlad New Member

    Messages:
    3
    Likes Received:
    0
    GPU:
    eVGA 6800 Ultra @ 450/1200
    Countless people have tested this and done it right including myself with a probe mounted on the GPU and see an increase from driver to driver.

    We aren't crazy, just because you don't get it or see it doesn't mean it doesn't exist.
     
  17. Mr_ALLroy

    Mr_ALLroy Ancient Guru

    Messages:
    1,994
    Likes Received:
    19
    GPU:
    EVGA RTX 3080 FTW3
    Nuff said!? Thanks for the work pumpkinlad
     
  18. Grendel_66

    Grendel_66 Master Guru

    Messages:
    793
    Likes Received:
    0
    GPU:
    MSI GTX 970 Gaming 4G
    The GPU is a fairly big chip that heats up unevenly. It's thermal diode is probably located in an area that heats up the most during operation. The location you stuck the probe in may not be affected by the effect you try to disprove, If you can't reach the chip I'd suggest to stick the probe to the heatsink -- it's the most likely place for more heat energy to go :)
     
    Last edited: Mar 11, 2005
  19. Emanon75

    Emanon75 Guest


    Well, yeah.
    As I said, it was done quickly.
    Just kinda shoved it in there..

    Had nothing to do w/ new/old drivers & temps.
    Just wanted to see if the cp was remotely close. ;)

    Admittedly, the fact that I saw no increase could be due to the way I tested. Could be due to the fact that my card's flashed, or it could be that some bios revisions or certain cards experience a rise, some do not.. I dunno.

    But It's still a driver issue.. :p
     
  20. 911medic

    911medic Master Guru

    Messages:
    211
    Likes Received:
    0
    GPU:
    BFG 6800GT OC (@400/1000, NV5 Cooled)
    I assume then, you can explain why if it's a simple 'reporting' bug (and not a true temp increase), when you apply that fix the temp slowly falls back to normal over a period of time instead of instantaneously displaying the 'correct' temp? Could it be that the temp increase was real and that the HSF is cooling the card back to normal? :rolleyes:
     

Share This Page