6800GS and 8I945PL-G crash processing graphics

By oRB06 Ā· 33 replies
Apr 26, 2006
  1. I need some serious HELP and this is going to be quite a long post so brace yourself.
    The Specs

    System Info Version: 2.2
    Operating System: Microsoft Windows XP
    Processor: Intel PentiumĀ® 4 @ 3.42 GHz
    Graphics Card: NVIDIA GeForce 6800 GS
    Total Physical Memory: 1 GB
    Free Physical Memory: 691.05 MB
    Motherboard Manufacturer: Gigabyte Technology Co., Ltd.
    Motherboard Model: 8I945PL-G
    Description: Primary Display Driver
    Manufacturer: NVIDIA
    Name: NVIDIA GeForce 6800 GS
    Total Local Video Memory: 512 MB
    Display Driver: nv4_disp.dll
    Display Driver Version:
    Driver WHQL Certified: Yes
    DirectX Version: 9.0
    Total Physical Memory: 1 GB
    Free Physical Memory: 691.05 MB
    Total Pagefile Memory: 2.40 GB
    Free Pagefile Memory: 2.16 GB
    Operating System: Microsoft Windows XP
    Version: 5.1.2600
    Service Pack: Service Pack 1
    Locale: GB
    Desktop Width: 1280
    Desktop Height: 1024
    Desktop BPP: 32
    Name: WDC WD1600JD-00GBB0
    Manufacturer: (Standard disk drives)
    Capacity: 149.05 GB
    Model: WDC WD1600JD-00GBB0
    Interface Type: IDE

    Ok, so theres the entire spec out of the way, I believe this rig is using a 350w power supply, although dont quote me on that.
    The Problem
    Right, I bought all the components in this pc, apart from the HDD and the case, abpout three weeks ago, which makes the HDD and case less than a year old. This rig ran fine for about 5/6 days, and then when i was playing Guild Wars one day, it crashed. I got a black screen, and it just sat there, no reboot, no error message, nothing. I manually rebooted and tried again, this time the PC lasted even less time, and died after about 4 minutes of graphics. Curious, I ran another game, Jedi Academy, much less intensive, this ran for about 10 minutes, then died. Seriously worried now, I turned to some stress testing freeware to asses the problem. Here are the results.

    I used Heavyload to stress the CPU for 30 minutes, the CPU load was 80%-100%. The temprature as calculated by Everest stayed at around 49, 50 degrees throughout, there were no crashes.
    Heavyload again, I allocated ALL of the ramspace that i have, until i had none left. There were no crashes.
    I ran both the CPU and RAM tests simultaneously, and there were no crashes.
    I downloaded and ran 'Video Stability test', by freestone group. After 5 minutes of running, the whole PC bombed out to black, and sat there.

    I restarted the PC, now convinced that I had nailed the problem on the graphics card somewhow. I ran the test again, looking closely at the temprature. Strangely, it didnt go above 55 degrees centigrade, so it wasnt overheating, sure enough after about 5 minutes it crashed out to black again.
    Applying logic here suggests that the GPU isnt actually malfunctioning because it DOES provide top notch graphics, and it DOESNT overheat. but for some reason, the whole system dies when you run some graphics for more than 6 minutes. The time isnt exact either, it varies, between 5 and 10 minutes sometimes.
    I went back and retested the CPU watching the temp carefully for a whole, again there was no crash and the CPU temp didnt tise above 53 degrees.
    I ran the GPU test and watched the CPU temp, again nothing above 53 degrees.
    So nothing is overheating, the GPU runs fine for a while, then dies and it doesnt appear to be the RAM. what is going on here.
    After extensive testing the PC finally threw me an error screen,
    Ive seen this about twice during the testing of this rig, and the last time it happened, (yesterday) my PC was irrecoverable. I had to format the HDD and reinstall windows, which is the system im talking to you from now.

    Honestly Have no idea what the problem is, i cant seem to narrow it down.

    Be aware though, the bios did thow up some interesting notes such as

    "OC FAILED please reset OC in BIOS" - i wasnt overclocking anything
    There were several times after crashes where the bot plain failed to load, or even locked up part way through, most of the time it just refused to even show me the bios.

    6 days plain sailing, then this, any help would be greatly appreciated as im about to go f**king crazy.

    Many Thanks
  2. es84

    es84 TS Rookie Posts: 48

    BSOD IRQL_NOT_LESS_OR_EQUAL is a windos issue , howeeve this can be a n issue with the ram as well . testing the ram is not always accurate , i have come accross many case where the test on the ram passed the test but the system gave bsod with it , it can be with the drivers of the video card, this is not a temperature issue.
    YOu need to set the virtuaL Memory to system managed .And remove all geforce sfotware and drivers. I believe the mobo has a built in video card , ty using that , there wont be any issues.
    in case the issue is still therfe frst
    then run the system on 1 ram card and then take it out and then run on the other .
    this will npot only issolate the issue fo us , we will fix this as we;;ll

    ( sorry for the speelings need speeling lessons
  3. KingCody

    KingCody TS Evangelist Posts: 992   +8

    hi oRB06, welcome to techspot :)

    well you've certainly done your homework.. lol

    since you've pretty much tested everything else, and they all work fine... the first thing I would look at is the power supply.

    since you don't even know the wattage, it's probably a garbage generic brand PSU. please list your current PSU specs (brand name, total wattage, and list the max amperage of the +12v line)

    if you have another PSU or have access to one, try swapping it.
  4. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Ok, im busy staring at the sticker on the side of my PSU, which is an Asus - A36F...it says here that it gives out a combined wattage of 340W. I dont have access to another PSU and in response to es84, i have no monitor that will plug into this mobo other than the one im using, which only plugs into the Graphics card. So im stuffed for taking out the card and testing again.

    Note: This PC works fine in windows, i can throw Photoshop CS2 at it, and run itnes messenger and several other proigrams without it even twitching. Yet 5 minutes og graphics will kill it...

  5. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    PS - heres a pic of my PSU sticker. You can probably make more sense of it than me.

    *Link Dead*

  6. KingCody

    KingCody TS Evangelist Posts: 992   +8

    the pic link you provided doesn't work, it just goes to a generic GMAIL page asking to sign in.

    i wasn't even aware that asus made PSUs, they are a good name for mobos, but PSUs???

    anyways, the reason it looks like a PSU issue is because it only crashes/shuts off during graphic intensive apps. modern graphics cards use alot of power especially when they are working hard.
  7. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Sry About the Link. True, they do, this particular card uses two of the standard power inputs that a CDROM or HDD would take. But it still doesnt excuse that everything worked fine for nearly a week without fault. Why the sudden change..?

  8. KingCody

    KingCody TS Evangelist Posts: 992   +8

    PSUs power output decreases over time, and can be greatly reduced by heat and dust.

    Unless somebody else chimes in with a different theory... I'd say you have a dying or over-taxed power supply.

    buy a new PSU, and if it doesn't work, then just return it (although i would suggest replacing that PSU anyways) get a good brand like Antec, Enermax, Sparkle, or Fortron rated at 350w or higher.

    a quality brand rated at 350w will produce more usable power than a generic 500w. the most important factor is the amperage on the 12v line
  9. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Looking at the results i have, Id say you're probably right. The other piece of supporting evidence is that when i leave my Pc off for say...5 hours, then use it with grpahics, it lasts anything up to 15 minutes, rebot, then 7, reboot then 5, reboot 1.30 and finally it dies altogether. I'm planning on buying a new PSU now, I cant afford anything too pricey Id say something in the region 50-70 quid? can anyone suggest anything?

  10. KingCody

    KingCody TS Evangelist Posts: 992   +8

    you can afford a very nice PSU with that budget.

    just as an example of a quality PSU... I have this one ,you can buy this line in higher wattages too and they don't cost much more. The only reason i'd suggest this PSU other than the fact that I own it (lol :D) is that it's almost silent, your HSF and HDD will be louder.

    I know you can't buy from newegg outside of the US, i was simply using it as a reference. the price on newegg is $60, which if I have done my conversions correctly would be about 33 "quid", well below your budget.

    so there's my suggestion, but any of the brands i mentioned earlier are also reputible PSU brands.

    good luck :)
  11. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Cheers Cody, Ill go shop for a new PSU and repost to let you know if that solves things (probably will)

    Many Thanks
  12. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Ok, before I go buy this PSU, theres one more thing...when i boot, occasionally the bios will freeze for a few seconds, this happens when it is running the line
    As this is related to RAM, this reopens my suspicion that RAM coulod be the cause? More tests?

  13. KingCody

    KingCody TS Evangelist Posts: 992   +8

    couldn't hurt to try...

    download and run Memtest86 to check for errors (it's a free program, google for it, intsructions how to run it will be wherever you download it from)
  14. es84

    es84 TS Rookie Posts: 48

    this is not a psu issue , you need to run hard ware test, it would be help full if u could post the stop error , with the details . and did u try to run the system on the onboard video card
  15. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    es84, i have nothing that will plug into my mobo so that I can view it on a screen, nothing looks compatible. I ran a series of hardware tests as you suggested. Heavyload ran RAM and CPU tests simultaneously for a good hour with no problems. THe problem i am having always occurs very quickly.
    I cant decide between PSU and RAM as the culprit. AS I mentioned, the "memtest" line does very occasionally freeze up a little. But it seems more likely to me for it to be PSU.

    PS I have no other GPU i can swap out to test with. THe error message i get is very rare, 9/10 the PC gives me a black screen, and i have to reboot manually. the other 1/10 happens after a lot of consecutive lock ups.

  16. KingCody

    KingCody TS Evangelist Posts: 992   +8

    and you say that because???
  17. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    I went out and bought myself a 450w PSU, advised by my local computer store that the 213w PSU that I had was most likely the cause. I installed and booted with the new PSU, ran a game, and it crashed within around 7 minutes of starting. As you can imagine, im now slightly stressed.

  18. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    I just noticed, rather weirdly, that if i have my PC on and running, and press lightly on the case above where the PSU sits, there is a crackling sound on the speakers...this sounds pretty dodgy no?

  19. paranoid guy

    paranoid guy TS Rookie Posts: 420

  20. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Hi Paranoid, Yeah it did come with a cable like that, but the new PSU i bought has a plug that can go directly into the card, I am now using the PSU direct cable.

  21. paranoid guy

    paranoid guy TS Rookie Posts: 420

    PSU direct AND a 450w PSU! You should be sailing... but you're not. I'd take it up with the man upstairs at this point, and my pc would be suffering from a hammer wound, I admire your patience.
    If your pc crashed do you have any minidumps?
  22. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    My patience is wearing kinda thin right now... I dont know about the minidumps. I installed Guild Wars just now, logged in and as it loaded the first play area it began downloading new files and updates, still in full screen mode, this continued for around 5 minutes. It then froze, but instead of a black screen, i got a frozen windows screen with a message from GW saying "a seripus error has occurred". So thats the first evidence of it not freezing straight to black, and it kinda makes me feel like the program ran out of space for soemthing as it was working... who knows.
    Ive tested everything replaced the PSU formatted the HDDs...no solutions as yet.

  23. paranoid guy

    paranoid guy TS Rookie Posts: 420

    Oh man, I feel sorry for you. Like the others said, your RAM could be holding you back, your graphics card robs 512mb. Or, maybe the card is just...broken. If KingCody couldn't figure it out, I'm in WAY over my head. Sorry man :(
    But if you can post the minidumps, every new one you can find, maybe there's at least hope of diagnosing the problem, never mind solving it. This really saddens me, a totally new rig and all...
  24. oRB06

    oRB06 TS Rookie Topic Starter Posts: 30

    Paranoid, i dont know a lot about minidumps or how to get to them...where can i find them and how do i get them up to see whats causing the freezes?

  25. paranoid guy

    paranoid guy TS Rookie Posts: 420

    Whew, found it! Howard wrote this about minidumps:
    So, like he says, minidumps can be found in c://windows/minidumps...I think that's what he said. It's at the bottom of the post, in red. I thought of another question(probably stupid) Do you have all the newest drivers for your motherboard/graphics card?
Topic Status:
Not open for further replies.

Similar Topics

Add your comment to this article

You need to be a member to leave a comment. Join thousands of tech enthusiasts and participate.
TechSpot Account You may also...