Crash help...

[derekfvk]

Originally posted by: mechBgon
Troubleshooting starter

1. Overview of the problem
   Computer is randomly freeze/crashing, particular during games
   
2. Full description of the problem and symptoms
   When i am playing games, sometimes i get a random crash. The crash is like the game freezes, the sound starts repeating really fast, almost like a buzz. The only thing to do is hit the reset button. Once i hit reset, the post code indicates a video error, (ami bios, 1 long 2 short). But if i power off, then power on the post screen comes up and says "The previous performance of overclocking is failed" (even though i was not overclocking!!!) I then hit space bar to go into bios, save bios and exit, and everthing is fine.
   
   This is generally during games, it started with far cry 2, i thought it may have been an issue with the game, but then it showed up in crysis warhead, the witcher, and most recently arkham asylum among others. In arkham it seems to happen at the end of the combo when it goes into slow motion knocking down the last bad guy. But other than that i havn't found any patterns.
   
   it would also happen whenever i tried to update the same game with the ingame windows live update.
   
   
3. Did it work normally at one time, or has the problem always existed?
   this seems to have appeared after i changed processors, but ... read on
   
   
   
4. Is the problem consistent and repeatable, or entirely random, or semi-random?
   
   see above
   
   
   
5. I already tried these steps:
   I am fairly confident it is not overheating, all system temps remain stable and system is well cooled.
   
   I have stress tessed the cpu with 4 windows of cpu burn in, for around a day long, with no issues, also stress tested with prime 95 for a long (approx 20 hours ) period with no issues and good heat dissipation, so it makes me wonder if it is the cpu?
   
   I have stress tested the video card with ati tool, video card stress test, furmark ozone3d, and 3dmark 06 with no issues. furmark raised the temp way higher than any game has and nothing wrong.
   
   I updated motherboard bios thinking maybe it was an issue that had to be straightened out with the chipset, but no difference. (before i update i had ami v1.8, and instead of 1 long 2 short, it was 8 short beeps, but both are video card related)
   
   all drivers are up to date..
   
   tested memory for 20 + hours with memtest 86+ and windows memory diagnostic, both reported no errors.
   
   
   
   
6. My software:
   • windows xp, 5.1 with service pack 3
   • antivirus and firewall are comodo internet security
   • Firewall
   • Feel free to add more software details that you think might be relevant to the issue
   
   
   
7. My hardware
   • System brand & model, if it?s a pre-built system such as a Dell, HP, Gateway, etc. If it?s not, fill in as much of the information below as you can.
     
   • Motherboard brand & model- MSI k9a2 neo-r
   • CPU model (and is it overclocked?)- amd x4 940 be (at stock)
   • Video card(s) evga 8800 gt, 512mb
   • RAM modules brand & model patriot ddr2 800, 2x 2048
   • Power supply brand & model silverstone st50ef plus 500 watt
   • Feel free to add more hardware details that you think might be relevant to the issue- also have sound blaster audigy 2 zs....
   
   
8. Other information that might be relevant
   (any other information that might be related to the problem)
   i can provide any more info you might think of...

So i just wanted to hear some opinions before i get in there and swap processors again, it did start when i changed to the 940, from 5000 be, but as i said earlier the cpu was stable in tests, and good temps, plus the bios beeps for video made me wonder, so i just want some second opinions, thanks.

 

[mpilchfamily]

What CPU did you have and what did you upgrade too? This may be a problem with the chipset not handling the new CPU correctly so a bios update may be needed. First thing to do is pop the old CPU back in. If problems persist you know its not the CPU. Then try rolling back the video card drivers if you have already reinstalled the most recent ones.

 

[derekfvk]

I had amd x2 5000 BE, swapped for x4 940 BE,

and as i said, bios is up to date,

 

[Intexity]

if i am reading the specs right you have an 8800gt on a 500 watt psu? unsure what the draw is on the amd but think you would be happier with a slightly heftier psu. just my two cents. good luck

 

[derekfvk]

i posted over at msi forum and someone there mentioned psu as well, the thought had crossed my mind but i kind of dismissed it since the other processor had no issues, but since it is a lower wattage i may try that route first.

and it makes more sense since i added another hard drive at the same time

what amperage should i be looking for on the 12v rails? something around 60 combined?

at the same time however, silverstone are fairly high quality units, so is it odd that a cpu change could do this?

 

[derekfvk]

well, I ran prime95(4 threads), large ffts, round off checking enabled, along with furmark, and It crashed within 5 minutes, so something about when cpu and gpu running together.

 

[ekoostik]

Did you rule out overheating in your latest round of testing? What are you using for a heatsink for the CPU? Besides the cpu that's the other component that changed. Are you sure it was re-installed correctly and makes enough contact with the chip?

 

[derekfvk]

Ok, so I think i have made some progress, it doesnt seem to be the cpu.. but it is the video card. I did a few more stress tests and noticed that the card (using furmark) was reaching 100+ degrees C (while cpu stayed in mid 30's). Around 105, boom, computer crash. I didn't think I was reaching this high of temps in games though, so I need to find a way to monitor the gpu temp in games. It is kind of weird that this started with the new cpu though.... but i put in a new cpu fan as well, I added a xigmatech hdt -s1283, so maybe it changed airflow or something? I need to explore this a little further tomorrow. Will also apply some mx -1 paste and reseat the gpu heatsink. and improve case airflow if possible.

This is kind of puzzling still, because from alt-tabbing in and out of games, or looking at evga precision after a session, the highest gpu temps are typicall less than 80 under load in games.

 

[ekoostik]

Run a program like GPU-Z that lets you log the GPU temperatures to a file, you can load into Excel after a gaming session or after a crash and see what the temps were doing.
http://www.techpowerup.com/gpuz/

 

[derekfvk]

So my cpu cooler was facing down (fan on top blowing into heatsink, towards bottom aka vid card), so I took it out, turned it around so it blows up, away from other components (maybe better flow to exhaust fans as well..) and reapplied some mx-2 paste, took out the 8800 gt, took the cover off the heat sink and gave it a good cleaning (dust bunnies in there) and now my gpu temps are running 20 C lower under load, mid sixties is the highest it reached in gameplay (mid 70 under furmark torture test), cpu is a few degrees cooler as well! So this seems to have solved my problem!

Thanks for all your input!