Jump to content


 


Register a free account to unlock additional features at BleepingComputer.com
Welcome to BleepingComputer, a free community where people like yourself come together to discuss and learn how to use their computers. Using the site is easy and fun. As a guest, you can browse and view the various discussions in the forums, but can not create a new topic or reply to an existing one unless you are logged in. Other benefits of registering an account are subscribing to topics and forums, creating a blog, and having no ads shown anywhere on the site.


Click here to Register a free account now! or read our Welcome Guide to learn how to use this site.

Photo

is my gpu or psu or mobo failing? (this is a weird one)


  • Please log in to reply
34 replies to this topic

#1 bluebolt

bluebolt

  • Members
  • 23 posts
  • OFFLINE
  •  
  • Local time:02:33 AM

Posted 26 December 2013 - 01:20 PM

windows 7 64 bit
gigabyte z68p-ds3
8 gig gskill ripjaw ram (2 dimms)
i5 2500k
gtx 560
650 watt (not budget ps)

It might be impossible to prove without me having a spare mobo and psu, but im curious of your opinions :)

System was ok for over a year but then I had my display driver randomly failing and recovering, regardless of what load it was under. I could be pushing the system quite hard and it would be fine though!

Eventually my system became unbootable after weird bsods.

I checked the memory overnight with memtest86+  no problem there, tried different harddrives and isntallations same problem.

If I tried to isntall drivers for the gtx 560 or for some reason the integrated graphics the sytem would become unbootable! If I took the card out, i could boot up and the integrated graphics would work for the system. I tried my other pci slot too, same problem.

I put my old 6600gt into the system, stress tested it with furmark till it hit about 80c. i was able to use the system stably with it for days :)
Although if I tried the windows experience index i would get a bsod when it tried to do windows media decoding? I didnt bother updating the os for other installations,  maybe the old card needed os updates for some of the tests? ill update later to check.

I replaced the 6600 with the 560 again and immediately the system turned itself off for a few seconds... at the post screen It said there has been boot failures because of overclocking/ changes of voltages.... and the current bios readings may not reflect the real ones.

If I loaded up on a drive with graphics drivers installed i would get a bsod upon hitting the desktop
if i loaded up on one without gpu drivers i could get to the desktop running on the basic windows vga driver.

Any ideas that can increase the chances of us guessing which component is failing?
Do i have to have drivers before I can underclok my gpu?

Please help me solve the mystery,
Thanks!!
 



BC AdBot (Login to Remove)

 


#2 mjd420nova

mjd420nova

  • Members
  • 1,715 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:11:33 PM

Posted 26 December 2013 - 01:31 PM

The BSOD with the older good card should be expected, the OS was characterized with the other card.  One thing i'd try would be if you have the original installation disk to do a repair with the new board installed.  That would allow the OS to find the most recent drivers for the card and things may return to normal.  Either a driver has become overwritten by another program.



#3 bluebolt

bluebolt
  • Topic Starter

  • Members
  • 23 posts
  • OFFLINE
  •  
  • Local time:02:33 AM

Posted 26 December 2013 - 01:37 PM

Thanks, I have tried repair installations unfortunately to no avail.

I have aslo tried several fresh installations after formatting the hd's. On one clean install the first thing I installed was the gpu driver for my 560. it still messed up,

bsod on reboot


Edited by bluebolt, 26 December 2013 - 01:38 PM.


#4 mjd420nova

mjd420nova

  • Members
  • 1,715 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:11:33 PM

Posted 26 December 2013 - 07:41 PM

Last time one failed like that in my experience, it was an IDE control chip on the MOBO.  It would intermittently write trash to the drive during repeated install attempts after repairs failed.  Hopefully it's not that serious.  If you have the disk, a new repair board should install with ease.



#5 zingo156

zingo156

  • BC Advisor
  • 3,333 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:01:33 AM

Posted 10 January 2014 - 01:45 PM

Sounds like the video card to me, or possibly a driver related to the video card. I have seen many failing gpus which would work fine with the standard windows vga driver. In safe mode this is default so most of the failing cards that I had seen would work fine in safe mode but crash to normal mode when the actual video card supplier driver was loaded. I would try a new driver or an older driver for the video card (from archive).

 

Some things I noticed with failing gpus throughout the years: most of them displayed funny colors or strange patterns before crashing (not all). Most of the crashes seemed to happen more frequently when the card was heated up (not all).

 

If a different video card works fine in the same pci-e slot I would highly assume that it is either the video card itself (hardware) or the drivers for the card.


If I am helping you with a problem and I have not responded within 48 hours please send me a PM.

#6 bluebolt

bluebolt
  • Topic Starter

  • Members
  • 23 posts
  • OFFLINE
  •  
  • Local time:02:33 AM

Posted 04 February 2014 - 11:03 AM

Bad news everybody!

 

I rma'd the gpu and there was nothing wrong with it. So Unlukcy! I am tempted to just buy a new bloody psu, but it could be the mobo. Why would the rpescence of my gpu make it impossible to install drivers for either the discrete OR onboard graphics. Tried many different drivers..  ,my 6600gt worked fine when stresstested. cant even get the drivers installed for the 560.

the 6600 will pull more when stressed than the 560 would at idle, so i fear it is not the psu..



#7 zingo156

zingo156

  • BC Advisor
  • 3,333 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:01:33 AM

Posted 04 February 2014 - 11:51 AM

Download and install:

WhoCrashed 5.00
Comprehensible crash dump analysis tool
for Windows 8.1/8/7/Vista/XP/2012/2008/2003 (x86 and x64)

After installing click the analyze button in the top left corner, copy and paste the results here. I am wondering if the drivers for the 560 are having a conflict with some other driver.


If I am helping you with a problem and I have not responded within 48 hours please send me a PM.

#8 zingo156

zingo156

  • BC Advisor
  • 3,333 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:01:33 AM

Posted 11 February 2014 - 09:53 AM

It might be an issue with one of the lanes related to pcie slots, this could be a problem with the main board or the cpu, it is possible that the 6600 makes use of different lanes or less lanes than the 560 and this is why it works fine but the 560 crashes. I would recommend trying a different main board or cpu if you have the option to do so.


If I am helping you with a problem and I have not responded within 48 hours please send me a PM.

#9 bluebolt

bluebolt
  • Topic Starter

  • Members
  • 23 posts
  • OFFLINE
  •  
  • Local time:02:33 AM

Posted 27 February 2014 - 04:43 PM

I just put it back in to the computer... and 0 problems with the integrated driver and the driver managed to install fine for the 560. i even stress tested it with heaven.. wtf.

 

I reckon they must have fixed it, but for somer reason didnt admit it? it was an oc edition that I dont think they can get hold of, perhaps they were supposed to replace instead of fix it. God knows. Thank goodness this bizzarre thing is over! thanks for the help.



#10 zingo156

zingo156

  • BC Advisor
  • 3,333 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:01:33 AM

Posted 28 February 2014 - 09:59 AM

You are welcome. I wonder if they dusted the card out entirely and it was heat related, or it is possible just re-seating the card a few times by pulling it out and putting it back in fixed the problem. I have seen this countless times with ram. Dust, moisture, etc, can build up and cause connection problems (reminds me of Nintendo days). Or maybe they did fix it and avoid the truth. Either way I am glad to hear it is working.


If I am helping you with a problem and I have not responded within 48 hours please send me a PM.

#11 bluebolt

bluebolt
  • Topic Starter

  • Members
  • 23 posts
  • OFFLINE
  •  
  • Local time:02:33 AM

Posted 07 March 2014 - 03:53 PM

Unfortunately I seem to be back to square one I am having firefox crashing after the nvidia display driver requires, and sometimes it doesnt recover and I have to reboot. On one of the times I rebooted in such an event, the computer turned itself off after 5 seconds and stayed off for another 5 before turning back on. No Idea why it would do that, but all this happened before when I first started getting problems.

 

I have no Idea why it would work for a while and start messing up again. This will ndoubtedly lead to windows not loading with the mere prescence of the 560's drivers being installed.

I decided to stress test my cpu out of curiosity. After 5 mins of folding@home, max priority, temps were at 85C. After 3 minutes of the prime 95 torture test.. It hit 95C, before I got scared and stopped it.  (shouldnt the computer have shut itself down, or at least given a sound warning?) the Tj. max of the 2500k is 98C.  I checked the temps with core temperature and hwinfo 64, they both agreed. I am using the stock fan, and the chip is not overclocked. I applied the thermal past properly in the centre when I first applied it.

 

I wont be able to afford to take it to the repair shop until April now. If I get to the stage of bsod and other crashes, it will be interesting to see what whocrashed has to say.

I changed the pci latency to the maximum value.



#12 zingo156

zingo156

  • BC Advisor
  • 3,333 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:01:33 AM

Posted 07 March 2014 - 04:42 PM

Sometimes those stock coolers can come lose over time. It would be worth taking it off, cleaning up the old thermal paste and then putting new paste on and re-mounting the heat sync. That cpu temp is way to high! Intel directly listed the max temp in Celcius as 72 for that cpu. If you truly had 98C then you likely have a heat related issue.

 

Intel 2500k http://ark.intel.com/products/52210/intel-core-i5-2500k-processor-6m-cache-up-to-3_70-ghz

 

Max CPU Configuration 1 TCASE 72.6°C Package Size 37.5mm x 37.5mm Sockets Supported LGA1155 Low Halogen Options Available See MDDS
If I am helping you with a problem and I have not responded within 48 hours please send me a PM.

#13 Datcoolguy

Datcoolguy

  • Members
  • 303 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:03:33 AM

Posted 09 March 2014 - 12:13 AM

You do know that most of Intel stock coolers already come with a pre-applied thermal paste right?

 

Anyways, make sure the cooler is properly seated in place


"If you don't understand how your computer works, you shouldn't be messing with it!"


#14 zingo156

zingo156

  • BC Advisor
  • 3,333 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:01:33 AM

Posted 10 March 2014 - 07:39 AM

I am fully aware they come with thermal paste, however if the cooler has come lose, or 1 of the feet that hold it on have broken, after removing it, you will need to clean both the cooler and the cpu and then re-apply thermal compound. 98C is still way too hot, double check your temperatures and if they remain high, check on your cpu cooler, it may have come lose.


If I am helping you with a problem and I have not responded within 48 hours please send me a PM.

#15 bluebolt

bluebolt
  • Topic Starter

  • Members
  • 23 posts
  • OFFLINE
  •  
  • Local time:02:33 AM

Posted 11 March 2014 - 02:41 PM

Thanks, I will reseat it and apply some more thermal paste.

 

Interestingly since I have altered the pci latency, the nvidia driver has recovered every time without me having to reset the computer.

 

I have been closely observing the temperatures of my gpu and cpu during the times the display driver glitched and recovered and they have never been high, so it could be a seperate problem. I was gaming a lot last year, and it is possible that the cpu was damaged permanently back then though.

 

I noticed the driver crashed a lot more when using firefox, andd about ten times as often when watching videos on youtube, especially when switching to fullscreen.

 

The thing I cannot comprehend is how the system became unbootable and remained unbootable on fresh windows instalations before I sent the gpu back. It should have immediately made my system unbootable when I put it back in recently, I do not know what changed. Is it possible that the bios stores some configuration details for the motherboards and/or motherboards voltages, and these were reset after my gpu was absent for a set amount of time.

 

This is still so bizarre, at least it always recovers after crashes now.  I think yhe cpu and mobo are both still under warranty too, so I will take to the repair shop next month.






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users