Jump to content


PSU, PCIe slot, or GPU (again)


  • Please log in to reply
3 replies to this topic

AliceUnchained #1 Posted 24 January 2019 - 08:29 PM

    General

  • Player
  • 38888 battles
  • 9,267
  • [322] 322
  • Member since:
    10-18-2011

As there are several knowledgeable people around, I figured it's worth a shot. 

 

Some time ago my GTX 980 Ti failed (started with black screens, which after analyzing dumps turned out to be TDR related crashes). Eventually after yet another reboot the Win OS logon just showed artifacts and remained stuck there. I replaced it with a 780 Ti I sill had lying around and everything booted just fine once more. The 980 Ti was RMA'ed (still within warranty) and I was given a full refund after some weeks. 

 

Now recently the 780 Ti appears to be showing similar behavior, issues. At first I thought the display was crapping out but it's working fine on the iGPU. However, the 780 Ti is also showing some peculiar issues I have not yet experienced before: sometimes trying to reboot the fans do not spin up (yet everything is properly connected, and power on LEDs are visible where they should be) OR after a forced reboot the card LED showed full load and the fans were spinning at maximum rotation. Even after a full discharge (PSU off, power cable disconnected, holding power button for extended period of time, removing 8-pin power connectors, etc) and rebooting the card it did the same, and once again after moving it to a different PCIe slot (and full discharge). After several failed attempts I just left the 8-pin connectors out and used iGPU. Coming home today system booted normally again on the GPU, and all is working fine for now.

 

Now given the issues with lack of power, and full power on the card on a reboot (after the initial crash has occurred) leads me to believe it's some power delivery issue. Which then could indicate PSU, PCIe slot, or just being unlucky and having another card crap out. Has anyone else ever experienced this curious behavior?

 

For the sake of completion:

Win7 64-bit Ultimate

Intel i5 2500K @ 4,6 GHz

2 * 8 GB Kingston HyperX @ 1866 MHz

Asus GTX 780 Ti Matrix Platinum (previously an Asus 980 Ti Matrix Platinum) @ 1006/1072 MHz (default factory OC)

2 * OCZ Vector180 240 GB (OS and games, including WoT)

2 * WD Re 5 TB

Asus Xonar D2X (PCI)

Seasonic X-Series 850W Gold

 

N.B. Overheating is not the issue (been monitoring that as well, even at low/moderate fan speed it does not exceed 80º ). Also, please spare me any driver suggestions. I simply would like to know if anyone ever experienced the odd fan/power behavior on reboot, and/or recognized it as a specific hardware issue. 

 

Edit: Using iGPU I've discovered that both when GPU fans do not spin, and spin at full speed the card is not recognized in Win OS.


Edited by AliceUnchained, 31 January 2019 - 09:47 PM.


AliceUnchained #2 Posted 31 January 2019 - 09:46 PM

    General

  • Player
  • 38888 battles
  • 9,267
  • [322] 322
  • Member since:
    10-18-2011

94 views yet no one had anything useful to add? Well that is disappointing. Since the previous post I have replaced the PSU PCIe power connector (2x 6+2 pin) and have not had any issues. But then again I've also barely played WoT or any other games so minimal load for the most part.

 

Oh well, already have a new CPU, GPU, case, PSU, RAM, surge protector and power cables to build a new system from scratch. Only thing left is deciding on a motherboard, most likely will be Asus Maximus XI Hero. Or Gigabyte Aorus Master.



jack_timber #3 Posted 01 February 2019 - 10:47 AM

    Major

  • Player
  • 35819 battles
  • 2,535
  • Member since:
    07-26-2014

Well Alice hardware problems are not easy to diagnose on a forum, can be difficult and obscure even it's in front of you apart from swapping 'bits' out.

Anyways GL with the build....



AliceUnchained #4 Posted 01 February 2019 - 07:44 PM

    General

  • Player
  • 38888 battles
  • 9,267
  • [322] 322
  • Member since:
    10-18-2011
Oh I am very much aware of that. It's not that I am expecting anyone to come up with a detailed root cause analysis, or be able to provide with great accuracy the failing component. I was merely hoping some people would be able to provide some insights, based on having experienced similar symptoms. Internet search only led to some 'issue resolved after RMA' but I am still not convinced it's the GPU itself failing.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users