BSOD Under Normal Use, Possibly Gfx Card

I've experienced 2 BSOD's today, both within 30 minutes of turning on the computer. I was doing nothing extraordinary either time, just browsing the web (in fact, the second time I was getting ready to post this). The blue screen flashed too quickly but I think it was stop error 116 and the file was ati something .sys. I suspect gfx card because of this. I've attached all the dump information.

OS is OEM, reinstalled 1 year ago. Mobo, CPU, RAM, secondary HDD are 2.5 years old. SSD (boot drive) is 1 year old. Gfx card, PSU are 6 months old.

Other maybe useful info about the gfx card is that it has been behaving strangely before. Last week and other times, Windows Aero would suddenly turn off, and the gfx temperature would rise ~5-10C and fan would go crazy. Rebooting fixed this, at least temporarily.

"It's not a true crash, in the sense that the bluescreen was initiated only because the combination of video driver and video hardware was being unresponsive, and not because of any synchronous processing exception".

Since Vista, the "Timeout Detection and Recovery" (TDR) components of the OS video subsystem have been capable of doing some truly impressive things to try to recover from issues which would have caused earlier OSs like XP to crash.

As a last resort, the TDR subsystem sends the video driver a "please restart yourself now!" command and waits a few seconds.

If there's no response, the OS concludes that the video driver/hardware combo has truly collapsed in a heap, and it fires off that stop 0x116 BSOD.

If playing with video driver versions hasn't helped, make sure the box is not overheating.

Try removing a side panel and aiming a big mains fan straight at the motherboard and GPU.

Run it like that for a few hours or days - long enough to ascertain whether cooler temperatures make a difference.

If so, it might be as simple as dust buildup and subsequently inadequate cooling.

I would download cpu-z and gpu-z (both free) and keep an eye on the video temps

Click to expand...

http://www.sevenforums.com/crash-lo...op-0x116-video_tdr_error-troubleshooting.html

Ok. I'll try that. I've always had problems with temperature with this card though. I have Speedfan, and the numbers it gives for CPU are 25-30C idle and 40-45C under load. With the GPU, it's idling at 45C and 60-65C under load. The lowest I've ever seen it idle is 40C. I have never gotten that idle temperature to go down. I've cleaned it best I could short of taking it out and taking it apart. I've tried messing with the fan speed controller with Speedfan, and I've set it so it's pretty much always at 50% or higher. I know the fan works because it's loud as f***. I've tried increasing the case fans to no avail.

It's got to be a temperature problem but I've haven't figured out why it idles so high.

EDIT: I found some dust on the fin things under the fan in the GPU. I've cleaned that out. I now also have large desk fan blowing into the computer as you suggested. The CPU is down a few degrees to 22-23C, but the GPU is still hovering around 45C.

Idle temperatures are not always accurate, so I would not worry so much about those. Your temperatures under load look good, so I do not think this is temperature related. Have you done the other steps in the 0x116 link provided by [URL="http://www.sevenforums.com/member.php?u=12080"]zigzag3143[/url]?

Besides the "blowing a fan into the computer" step, no. I will try Furmark, Memtest, etc. when I get home.

Also, I updated my drivers to the latest ones AMD released last week, version 12.4 I believe. Should I roll them back for the purposes of these tests?

Keep 12.4 on the system and see how things go. You can switch drivers if you continue to have problems. How has the system been running with 12.4? Any more stable?

So far so good with the 12.4 drivers. I also ran Furmark just now. It maxed at 80C and there was no instability, so that's good. I'll post something if I have more problems with the latest drivers.

By the way, you mentioned idle temperatures not always being accurate. Can you explain what you mean?

Most temperature sensor software is designed for measuring temperatures under stress. There is more calibration that needs to be done for idle temperatures because the idle temperatures are generally closer to room temperature and prone to error for that reason.

Also, it depends on the software that measures the temperatures. Some software will work better with certain systems but be horrible with other systems and vice versa. It is best to test multiple programs and see which gives a more accurate measurement for your system. Get a feel for what you would expect to happen as you stress the system and see if one monitor does a better job in those conditions.

In my experience, most monitors will be more accurate under stress than at idle, and I attribute this to the fact that they are designed to watch for high temperatures and not care as much about lower temperatures and also the room temperature error I mentioned earlier. Not sure if I answered the question properly since I am making educated guesses about what is likely happening. Given that your idle temperatures are quite a bit higher than room temperature, I would guess they are fairly accurate, but I cannot say for sure.

The stress temperatures are more important, at any rate, and if those are well below 70 C on your card, you are doing very well. My AMD HD 4850 in my desktop runs in the 90-95 C range under stress, for instance, and around 50-60 C under idle. It is clear of dust, but I have never been able to get it to run at lower temps. No problems with the system, though *knock on wood*.