CCC Keeps Crashing

[Erazor GTX]

Hey all,

I've been having this intermittent problem with my video card drivers lately, and I can't figure out if it's a hardware or a software problem. The AMD display driver will crash randomly, even if the computer is idling. Actually, It happens more when the computer is at idle than when I'm playing games, but I can't do one thing that consistently recreates the issue. After it crashes, I sometimes get artifacts on a small part of the screen, but not always. Plus, if I fire up a game, they'll disappear while I'm in game. My computer is set up to three monitors (1920x1200) and I'm running windows 8. I've already tried a clean install, but the problem came right back. I've also reinstalled CCC a few times to no avail. Is there anything else I can try? I just want to be as sure as I can that this isn't a software issue before I splurge on a new video card.

[Chrispy_]

Sounds to me like a faulty hardware - 90% likely to be the card but I wouldn't completely rule out PSU, motherboard or system RAM.

If it happens whilst idling, try running without drivers for a few hours - I mean the Microsoft "Standard Display Adapter". If it crashes then you have proof it's not the driver.
It will look like arse, but at least it's not the 640x480, 16-colour uselessness that it used to be.

[Erazor GTX]

If it happens whilst idling, try running without drivers for a few hours - I mean the Microsoft "Standard Display Adapter". If it crashes then you have proof it's not the driver.
It will look like arse, but at least it's not the 640x480, 16-colour uselessness that it used to be.

Thanks for the suggestion, I'll give it a shot and see what happens.

[Erazor GTX]

So I disabled the video card drivers, and no artifacts yet, but sometimes they can take days to show up. It's also been prime95 stable for about 14 hours now. I'll run memtest next, but I was thinking: Is there any sort of video card stress test that may help me out?

[Chrispy_]

I'm not sure how much will run without drivers but you should try furmark.

http://downloads.guru3d.com/FurMark-1.8 ... -2523.html

You said it happened more often when you were idle. Try running flash/silverlight/java web utilies for long periods - also set your sleep/hibernate/screen-off times much lower in case it's caused by power-saving functions.

[Erazor GTX]

I'm not sure how much will run without drivers but you should try furmark.

Thanks for the tip - How long should I let the test run? I'll reinstall the drivers if need be.

also set your sleep/hibernate/screen-off times much lower in case it's caused by power-saving functions.

When I first noticed the problem, I had my computer set to never sleep. However, when I reinstalled windows I kept the sleep timer at 30 min (because I thought the problem might be due to heat), and the problem still randomly happened.

Two other things I just thought of: I'm not sure if this is related or not, but I never had these problems on windows 7 before I upgraded - and none of the hardware has changed. However, the problems could also be related to hurricane sandy - my computer is connected to a UPS, but the power went in and out frequently when she hit our area. I really can't remember if the driver problems started around that time or not.

[vargis14]

I suggest doing/running Driver Sweeper.....you have to do it in safe mode but at least you are pretty much garanteed that you will remove all traces of old driver installs.

Here is a link to guru3d with instructions and a link to the driver sweeper program.
http://www.guru3d.com/content_page/guru ... eeper.html

I would give that a try....I have noticed even if you uninstall old drivers before installing new ones a lot of old driver junk can still be left around. If you just install new drivers over old drivers then I super recommend you do a driver sweeper run.

Good luck to you. V

I also recommend you install afterburner or some kind of overclocking software so you can monitor temps fan speed memory usage idle speed state and load and intermediate power states ...Or you can install gpuz that will let you monitor your voltage etc along with a little card stressing application that can also run a 3d deal in a little applet window to load your card.

[Chrispy_]

Thanks for the tip - How long should I let the test run? I'll reinstall the drivers if need be.

On a clean install of windows like you've done, you shouldn't need to mess with drivers, though I guess Vargis' driver sweeper suggestion won't do any harm.
Furmark includes a temperature monitor. In the unlikely event it's heat related you'll know but Furmark is probably going to get your card at peak temperature within 15 minutes.

As for the W8 driver/hurricane Sandy issue - you're doing the best thing first; Trying to reproduce the issue without drivers. That way you know if it's a config issue or a hardware fault.

[Erazor GTX]

So, I took it out of safe mode to run furmark (I couldn't get it to run otherwise ). It's going on about a half hour now with no problems. If the problem was with the power supply or motherboard, how would I troubleshoot it?

Also, I did a little bit of research, and found some other people having problems that might be related to mine? Maybe?

http://forums.guru3d.com/showthread.php?t=370988

Only difference it seems, is that my screen flashes black and I get the: "AMD Display Driver has stopped responding and recovered" message before I get similar looking artifacts on my screen.

[Chrispy_]

My only W8 machine is now an Nvidia GT650M, so I can't really help you further. Faulty power supplies and motherboards normally get worse under stress. If you are simultaneously running the CPU and GPU with 100% load (run furmark and Prime95 at the same time) for 20 minutes or so, it's probably not going to be a hardware fault.

My experience from troubleshooting issues before on W7/Vista is that AMD driver crashes are usually related to the graphics card being faulty. The one exception was a Firefox issue that turned out to be buggy GPU acceleration code that didn't manifest in Chrome or IE:

In Firefox 4 and above, Firefox Safe Mode also disables hardware acceleration, so the problem could be a video device or driver issue (see below). Try unchecking the option to use hardware acceleration in Firefox ("Tools -> Options -> Advanced / General").

I'm not saying that's your problem but it's one more thing to try if you've run out of other options.

[Erazor GTX]

Seems like the PSU is fine, so it look like I might have to replace my GPU in the near future. I just wanted to be sure that I had ruled as much else out as I could before I went ahead and splurged on another card. Thanks for the help, I really appreciate it!

[Starfalcon]

I had a similar one when my 6950 was dying, it started out with a random one every so often, until it eventually happened every 10-15 mins. After one massive crash and lockup, after a reboot, my card only showed half of the 2 gigs on the card for video memory. Within 2 weeks of that the card was completely dead and off to RMA.

[Erazor GTX]

It seems like it's happening 1-2 times every two weeks right now. I've got my eye on a 7850 to replace it. I figure I'll pull the trigger when the failures start to happen a little bit more frequently.

[Savyg]

I've had that problem with my 6870. It mostly stopped happening when I started manually running the fan at 50% before gaming sessions.

Seems like their fan code isn't pushing it up fast enough in the newer drivers.

(On another note I've had issues with games for a long time before I started doing this, and now they run flawless, so awesome.)

[Chrispy_]

I've had that problem with my 6870. It mostly stopped happening when I started manually running the fan at 50% before gaming sessions.

Seems like their fan code isn't pushing it up fast enough in the newer drivers.

(On another note I've had issues with games for a long time before I started doing this, and now they run flawless, so awesome.)

That sounds more like a fussy, on-the-edge card from the outset that you've managed to fix by improving the cooling.
"Fan code" is set by each card manufacturer in the GPU BIOS for AMD cards, not the drivers. You're thinking of Nvidia, most likely.....

[Savyg]

"Fan code" is set by each card manufacturer in the GPU BIOS for AMD cards, not the drivers. You're thinking of Nvidia, most likely.....

***,

That would be an issue related to hardware which is controlled by drivers.
As far as I know, there has not been a BIOS update for this card, but I can
check with our engineers. Have you tried installing our video card on
another computer to see if you get the same problem?

-----Original Message-----
From: ***
Sent: Thursday, January 12, 2012 10:01 PM
To: helpdesk@diamondmm.com
Subject: Diamond Customer Inquiry [ORIGINAL]

subject: 6870 fan speeds

comments: I`m becoming concerned that I never hear my 6870 fan power up
unless I set it to manual control and it`s becoming unstable in heavy gaming
sessions. Has there ever been a BIOS update for this product, or would that
be entirely driver side?

Sounds like drivers to me.

[ryko]

i kept having ccc crash for me in game during cod: black ops 2. this is with a 7850 and the latest catalyst version. finally figured out that turning on vsync in game would eliminate the crashing. it was like my system was running too fast and then crashing to desktop. capping the frames at 60 did the trick for me. no other games exhibited this problem though.....so YMMV

[Chrispy_]

Sounds like drivers to me.

There is a driver element which tells the card to use a particular state as defined in the BIOS, but I have been editing ATI/AMD graphics card BIOSes for over a decade now. The fan controls, voltages and clockspeeds for each state are determined by the BIOS. The drivers can give more information to the card about what state to be in (2D desktop, 2D video, 3D windowed, 3D fullscreen) but ultimately the fan speeds are based on state in the BIOS and these states are hardcoded by the manufacturer, per card. It's necessary since two almost identical boards from the same vendor can have different coolers. One requiring a 2000RPM blower and another requiring two 1000RPM 92mm open fans.

The drivers are generic and apply to thousands, if not tens of thousands of variants of Radeon from multiple manufacturers across several generations. The fan speeds for one model are likely (if not definitely) wrong for a different model. I think you're misunderstanding your response from Diamond Multimedia which is saying that your BIOS/Card is faulty and that a particular state including clocks, fanspeeds and voltages is not responding to a driver state correctly. That's still a hardware fault and your workaround is fine because it's avoiding this state in which your card is not functioning correctly.

[Savyg]

I'm guessing its just dust buildup in the card, since I'm not really going to remove the shield to find out I'll just have to go with that assumption.

[lilbuddhaman]

I had this problem with one of my CF'd cards not seated properly (it's own weight leaning out of the slot).