

Graphics Drivers Crash, Fail To Recover
#1
Posted 05 May 2013 - 01:32 PM
Windows 8 64 bit
AMD 6100 stock speeds
AMD Radeon HD 7770 stock speeds
8 gigs Corsair vengeance
Top 650W power supply
Any help would be greatly appreciated.
#2
Posted 05 May 2013 - 02:06 PM
If your computer has a detailed record of the crash, it will be there. I'm a little concerned about that power supply. I've never even heard of "top" (do you mean Topower?). It's possible that a bad PSU is causing bad power to the card.
Try going through a full run of 3Dmark11, or the new 3DMark (particularly the Firestrike test), or Unigine heaven, something that really runs the GPU at least reasonably to its limits, and see what happens. It's probably not a GPU hardware issue, so at least there's that.
#3
Posted 05 May 2013 - 02:16 PM
Catamount, on 05 May 2013 - 02:06 PM, said:
If your computer has a detailed record of the crash, it will be there. I'm a little concerned about that power supply. I've never even heard of "top" (do you mean Topower?). It's possible that a bad PSU is causing bad power to the card.
Try going through a full run of 3Dmark11, or the new 3DMark (particularly the Firestrike test), or Unigine heaven, something that really runs the GPU at least reasonably to its limits, and see what happens. It's probably not a GPU hardware issue, so at least there's that.
Yes I did mean toppower. its a brand new PSU (not even a week old) and it just replaced a corsair 450 (just because, had no issues with the corsair, really just getting older). Ill try 3dMark, and get back with the results.
Edit: no minidumps have been found.
Edited by Lagoth, 05 May 2013 - 02:33 PM.
#4
Posted 05 May 2013 - 04:54 PM
When I was last OCing my GPU, I would see that behavior. The GPU wasn't getting sufficient power, so the computer would just seize up.
Based on what I'm seeing, if I had to guess, I'd say your GPU is probably being starved by the PSU and crashing. It's a guess, but not a blind one. I'm guessing that it tends to happen in games that stress your GPU, while games that don't, like Skyrim and WoT, which aren't really GPU-killers; even FC3 really isn't that GPU-intensive. I know nothing of World of Warplanes (is it seriously called that?

Running a few 3D benchmarks should confirm or refute the hypothesis.
#5
Posted 05 May 2013 - 07:42 PM
#6
Posted 06 May 2013 - 03:11 AM
Catamount, on 05 May 2013 - 04:54 PM, said:
When I was last OCing my GPU, I would see that behavior. The GPU wasn't getting sufficient power, so the computer would just seize up.
Based on what I'm seeing, if I had to guess, I'd say your GPU is probably being starved by the PSU and crashing. It's a guess, but not a blind one. I'm guessing that it tends to happen in games that stress your GPU, while games that don't, like Skyrim and WoT, which aren't really GPU-killers; even FC3 really isn't that GPU-intensive. I know nothing of World of Warplanes (is it seriously called that?

Running a few 3D benchmarks should confirm or refute the hypothesis.
Ran all three 3DMark tests 3 times each and had no issues whatsoever. Temps hit 54 CPU and 52 GPU, and scored in what 3DMark says are valid results (whatever that is worth). And yea, it really IS call World of Warplanes

Bad Karma 308, on 05 May 2013 - 07:42 PM, said:
I am running Catalyst 13.4, drivers 12.104.0.0 (the most recent, I do believe). But the problem has been persistent for a while, spanning several driver updates, and game updates.
#7
Posted 06 May 2013 - 05:41 AM

Okay, I'm at a loss here for the moment, so I'll have to give it some thought. You could always step up to a more serious stress test, like doing a burn-in with Furmark for 5-10 minutes, just to be absolutely sure, but if your card is making it through the 3dmark Fire Strike test, it's probably okay.
Another thing you could try for MWO is switching it out of fullscreen mode, which is generally buggy. The Full Window setting is a good alternative, and I generally recommend making that change one way or the other (several people in my unit have reported far fewer CTDs).
#8
Posted 08 May 2013 - 01:40 AM
I've ran Furmark for about 20 minutes the other day, and caused no issues. Temps were showing 60-65 degrees.
#9
Posted 13 May 2013 - 03:38 PM
At random, in the middle of a match, my screen will freeze, flash black, and recover about ten seconds later (Problematic in a firefight, you might imagine.) It appears to be a driver failure but I don't even get a message for THAT.
My system packs a Radeon 7970 XFX Black Ed.
Running the VISION control center and watching the video performance shows my GPU capping at 100% usage immediately after the crash, even though it passively runs at about 70% at max graphics and 40% at the settings I have them at now. The GPU usage stabilizes a few seconds afterwards. (From what I understand, the freeze and flicker is a byproduct Windows trying to pick up whatever process the GPU lost so as to prevent a total system crash. At least, according to some guy with more tech knowledge than me.) My card runs at about 65 degrees at max setting and 55 at what I have right now.
I dropped my graphics settings to medium and turned off anti-aliasing, and that's reduced it from 2-5x a game to once every four or five games. It's still irritating as all hell though, and I'd like to see the game in all its glorious shininess.
EDIT: I also run the game in Full Window mode. That seems to make a difference too.
Edited by Phaeroth, 13 May 2013 - 03:39 PM.
#10
Posted 13 May 2013 - 04:20 PM
If I had to suspect, I'd say your PSU is failing to give sufficient voltage to your GPU. Download GPUZ and run it while putting your card under full load, see what your 12v rail does (for instance, mine drops from like 12.25 to like 12.15). Also look at your VDDC, which should be around 1.1v
#11
Posted 17 May 2013 - 02:05 PM
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users