PC crashing during gaming and prime95

dpbetter2000

Honorable
Jan 25, 2014
115
0
10,710
Hello everyone! First of all I apologize in advance due to my lack of knowledge about software and most diagnostics.

Here are my specs:
AMD FX 8350, never overclocked
ATI R9 280x 3gb vram
Asus m5a99fx pro r2.0 mobo
8gb corsair vengeance ram
Rosewill arc 750w PSU
Zalman cnps10x CPU cooler

So I started having these BSODs maybe 2 months ago, and at first I didn't pay it any mind, but it started getting more frequent. I started running diagnostics on each piece of hardware and so far nothing. I tested my hard drive, no corrupt files. I used memtest86 for 12 hours with no errors. I ran sfc/ scannow in the command prompt and no errors. I ran furmark and while my GPU temperatures got fairly high--around 65 degrees celcius--no BSOD.

Then I started prime95. On my first try with cpu fan speed at 50%, my PC had a BSOD (WHEA_UNCORRECTABLE_ERROR) after less than 10 minutes running the program. Next try, I put the fan speed back up to 100%, and it ran a little longer, long enough for me to get a fatal rounding error. Then while I was researching said error, I crashed again. Throughout this, my temps seem to have remained fairly stable, cpu temp is around 56-57 degrees celcius and core temp is around 40. I have 4 fans, 2 intake in the front, 1 exhaust in the back, and the CPU fan itself.

The only thing I can think of at this point is that my processor is failing, either that or overheating at temperatures that I believe should be completely fine.

PLEASE HELP, I'm happy to answer any other questions to the best of my ability
 
Solution
Your thermal margins are fine (they should be 10°C or higher); I'd RMA the CPU because it shouldn't fail that stress test. Even if the CPU was too hot, it should only throttle.

dpbetter2000

Honorable
Jan 25, 2014
115
0
10,710
I have not, in fact I don't think I knew that the motherboard itself could overheat. Any programs you would recommend for monitoring mobo temp?

As a side note, I'm considering buying a closed loop CPU water cooler--looking at the CoolerMaster 240m--in hopes that it will increase airflow (my current air cooler is HUGE) and keep everything a little cooler and quieter. Would this be a good investment?

 

dpbetter2000

Honorable
Jan 25, 2014
115
0
10,710
1 minute into the test, core 5 received a fatal rounding error, core usage goes down from 100%. Thermal margins at 23 celcius. 5 minutes in, core 6 receives the same fatal error and usage goes down from 100%. Thermal margins now at 20 celcius. At around 15 minutes, my thermal margins seem to stabilize at 15.5-16 celcius and no other cores have gotten a fatal error. Right as I'm about to stop the test and write this post--I'd say after 17 minutes of testing--my computer crashes with the error message SYSTEM_SERVICE_EXCEPTION.

EDIT: If it makes any difference, I have a bitfenix neos case, which I've heard is one of the worst airflow cases money can buy.

 

dpbetter2000

Honorable
Jan 25, 2014
115
0
10,710
Thank you! It's been about 2 years since I built this PC, any idea if AMD will actually replace it unless I have certain proof that it's the processor that failed? I don't want to go through the whole return process and take apart my PC only to have them tell me that they won't replace it.

 
You'll know if they'll replace it once AMD issues the RMA. Key in the required info at http://support.amd.com/en-us/warranty/rma and AMD should let you know within a few business days if you can return it. You should basically indicate the crashes in games, provide the Prime95 failures on 2 cores, the thermal margins reported by AMD Overdrive, etc. Basically the info that you provided in this thread, but they don't care about your PC case.
 

dpbetter2000

Honorable
Jan 25, 2014
115
0
10,710
Thanks again, hope this resolves the issue.