No clue on GPU problem

Toggle sidebar Toggle sidebar
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780
Hi!

I have this very weird problem with graphic cards on my pc as it tends to fail under pressure. I can game bf3 or run stress test with no artifact and really good framerates for a little while (usually 30 mins or something) with really good temps, when it suddenly BSOD's or gives me one form of an error message or another.

My parts:
* Gigabyte GA-Z87-DS3H mobo
* intel core i7-4770K @4.1GHz liquid cooled
* Cooler Master Seidon 240m
* Corsair Force 128GB SSD (Windows 7 Pro)
* Kingston HyperX 120GB SSD (OSX Mavericks)
* Gigabyte Radeon R9 280x Windforce OC edition GPU
* Crucial Ballistix 16GB Sport kit - RAM
* TP-link WLAN card
* Corsair CX750M PSU

This is my 4th card, RMA'd 3 GPU's. I can't imagine 4 faulty cards in a row, like - what's the chance?

Could it be software related?

Please help me with this, I've got no clue at all anymore.

Thanks!
Erik
 
Sort by date Sort by votes
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Sorry, its a 750w cx750m, tried 2 650W silver power ones plus this corsair one - all same problems
 
Upvote 0 Downvote
J

junkeymonkey

Polypheme
BANNED
Nov 11, 2013
17,389
4
66,960
sure been a lot of issues coming in to toms on these -R- card looks like ones to avoid right now [opinion]

try this --- uninstall the driver shut down- discharge the board remove the card from the slot and whip the ''golden'' fingers clean- insert the card and remove it and reinsert the card back to help insure it ''scratches in good contact in the slot hook it all back up and boot back to desktop - reinstall the driver and recheck..
 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Lol, didn't quite catch what you meant with "golden" fingers and all that, (a joke?:p)

Discharge as disconnect PSU from wall outlet and wait a few min or?
 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


;) I won't be able to test this until weekend unfortunately - but I'll try and report back to you ASAP I got results ;)
 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Yeah, I can do that - but do you mean after a BSOD or what? ;) - the TP-Link is a PCIe card, yes
 
Upvote 0 Downvote
J

johnbl

Polypheme
Nov 4, 2012
10,072
185
67,090
yes, you can post any memory .dmp files that you think are related to the current state of your machine.
(sometimes people post .dmp files that are several years old and unrelated to their problem so check the file dates)



 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Sure, but I've never done this before - could you please post a step by step guide or a link to a guide?
Thanks!;)
 
Upvote 0 Downvote
J

johnbl

Polypheme
Nov 4, 2012
10,072
185
67,090
generally the memory .dmp file will be located
c:\windows\minidumps

but the settings are user or OEM settable so it can be in other places.
you would search for *.dmp to find them on your drive

you would then get the file and upload it to a service like google docs or skydrive and make sure they have public access so it can be read. (up load procedures depend on what service you use)



 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Thanks;)

I'm a poweruser for PC so I got a good idea on upload - so no problem there. Just needed some info on the minidump. I found a few ones but I dont think they're related. It ran 75 mins stressing today perfect followed by 45 minutes of bf3 gaming - no problem, but opening a youtube video made the driver fail for some reason. I will upload minidump and tell you if I get one related ;)
 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Almost looks like Battlefield 3 works (failed a little, but no failing after I downloaded and ran DirectX again) - but Microsoft Flight Simulator X always ends up with an angry buzzing noise together with a freeze. No minidump up-to-date yet
 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Here we go, a minidump file:
https://www.dropbox.com/s/d9hl77j0dwc0lbq/080514-8876-01.dmp

Thanks SO much for help!;)
 
Upvote 0 Downvote
J

johnbl

Polypheme
Nov 4, 2012
10,072
185
67,090
thermal shutdown called by the BIOS in the GPU?
===========
only thing I can think of would be problems in the electronics caused by overclocks on the CPU, PCI bus or the actual GPU.
I would reduce all overclocks and see if the problem occurs. If not, then enable one overclock at a time until the problem happens.

my guess is the overclock on your GPU is causing some power/ heating issue for the GPU card. (you might want to make sure you have good air flow in your case)

- sorry, it is not much help






your bugcheck was an unknown code 0xA0000001 called by the AMD graphics driver
Image path: \SystemRoot\system32\DRIVERS\atikmdag.sys
Image name: atikmdag.sys
Timestamp: Thu Apr 17 19:13:16 2014 (53508A3C)




machine:
BIOS Release Date 01/20/2014
Manufacturer Gigabyte Technology Co., Ltd.
Product Name Z87-DS3H
Processor Manufacturer Intel
Processor ID c3060300fffbebbf
Processor Version Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz
Processor Voltage 8ch - 1.2V
External Clock 100MHz
Max Speed 7000MHz
Current Speed 4100MHz
memory:
Speed 1600MHz
Manufacturer 1315
Serial Number
Asset Tag Number
Part Number BLS8G3D1609DS1S00


 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


Thanks so much for answer. The GPU is clocked at 1100MHz core and that's stock as its the OC edition. It's my 4th GPU so I can't understand it being faulty again. My first startup on this PC corrupted my BIOS because of a cooler issue, but restored thanks to dualBIOS. This happened several times - can it be related? What do you think?
 
Upvote 0 Downvote
J

johnbl

Polypheme
Nov 4, 2012
10,072
185
67,090
looks like there is a BIOS update for your GPU
http://www.gigabyte.us/products/product-page.aspx?pid=4793&dl=1&RWD=0#bios

dated 6/3/2014 (confirm that i picked the correct card GV-R928XOC-3GD (rev. 1.0))


================
Note: the card runs at 1000Mhz but then automatically boost to 1100Mhz is there a way to block the automatic boost in the GPU?

- the bugcheck looks like it come directly from AMD (not windows) I would think they should know why it is being called.
(and I would think they call it because of a hardware issue in the Card, power fluctuation or thermal problem)

-Is there any chance that your BIOS slightly overclocks your PCI bus? A overclocked GPU on a overclocked bus sounds like a problem. (potential power problem or heat problem)
I know my old ASUS BIOS would set my PCI bus from 100MHz to 103MHz and that worked fine until I got a newer video card.

- is there a way to underclock your graphics card from the overclocked version to a standard clock for the GPU

- if the BIOS seems to get corrupted you might want to check for some auto overclocking set in the BIOS.




 
Upvote 0 Downvote
E

erikdude27

Honorable
Jun 24, 2013
266
0
10,780


This is what I know now:
- I changed a PCIe setting in the BIOS from auto to gen2 and that made the problems go away but I now got a slight lag now and then. A smoothness issue.
- My GPU is rev 2.0 and not 1.0 - still a VBIOS update?
- My BIOS may be overclocking the bus, but I don't know what setting to look for in that case.
- I could underclock using MSI Afterburner or similar but shouldn't the voltage need to be changed then as well?
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom