BSOD during stress testing on new computer

[omnage]

I recently built a PC with the following Specs:
MOBO8P67 Deluxe 3.0 1850 BIOS
CPU:Intel Core i7-2600K
RAM:G.SKILL Ripjaws X Series 8GB (2 x 4GB) 240-Pin SDRAM DDR3 1600 (PC3 12800)
Graphics cardsual EVGA GeForce GTX 570 HD Superclocked 1280 MB GDDR5 with SLI Link Enabled
PSU:Rosewill LIGHTNING 800W ATX12V/EPS12V
CPU cooler:Corsair Hydro Series H70 w/ 2 fans cooling liq
OS: Windows 7 Ultimate x64

I have experienced this before the stress test by playing games for several hours (6+) and at first I thought it was a heat issue. so I carefully began to watch the temps however my CPU never goes above 35c on load (19c idle), the video cards never seem to go above 50c, ram runs around 34c and mobo gives a temp of about 38c. no issue with heat.

I noticed that BCCode often referred to people coming back from hibernation or sleep however i can run the stress test on a clean boot and it will still fail.

BIOS is set to default setting with everything to auto. I have seen people talk about needing more power but i would assume even with dual video cards 800w power supply is adequate. is that possible?

This is a clean install from around 9/30 and all Drivers are up-to-date however i am using a beta bios according to asus. I moved up to it cause it says it improved system stability. Also i am using a beta driver from nvidia now as well, however before the update to this driver I was still getting this issue.

I am willing to give any info and am decently able to answer most questions about my build but this problem seems to escape me. Any help will be most graciously accepted. I am more than willing to experiment with settings and learn how to debug this issue if you have any guidance i would appreciate that as well.

See Attachments below for dump and PC health report.

 

[panais]

Hi omnage welcome to SF


You need to update your,


Marvell MV91xx AHCI/RAID software\driver

Marvell - Support



Finish with this steps and post back with any results.

 

[omnage]

currently running prime95 test again to see if i get the fail. was reading that it could be a ram voltage problem and not being high enough. adding like .2 bring it from 1.34 to 1.5 or something. will do further research on that as well. will know about the driver failing in a second or so when it decides to drop. takes 15 or so minutes.
Edit: Ok so while i was waiting for the enevitable fail i went searching for drivers from a non asus site and found some re-running test. will have new update soon.

 

[omnage]

Still nope help see attachment below. Dies without even showing a blue screen still.

 

[panais]

Did you followed the step in my first reply?


Do you overclock anything?

 

[omnage]

Yes current driver state for Marvell MV91xx AHCI/RAID driver is 1.2.0.1006.

No overclocking has been done. infact when i reinstalled windows i placed the bios back at default settings and ensured that nothing was overclock and it was set to normal/auto. i have yet to change anything.

 

[panais]

The minidump you uploaded is exact the same.

 

[omnage]

clearing logs and rerunning tests. will upload when done... not sure why it was exactly the same.. driver in device manager lists both sata and controller as 1.2.1006. otherwise same error is occurring as before.

 

[omnage]

Ok well let me know where you see them at so i can look (which file do you find driver version numbers, maybe i missed one) or i can post a png of my device manager showing driver info... but i deleted the logs and re ran the test again, still failed, generated new reports and zipped them together.

 

[omnage]

Here is where i stand. on this issue.
I have run the following test and they have all succeded
sandra CPU stress test - passed, no errors
Memtest86 - passed, no errors
prime95 small fft, passed for 22k no errors - ran for 5 hours

Next up is as follows
furmark
prime95 Inplace large ffts
sandra burn-in
- I do not expect to see an issue less it has something to do with power problems as i believe it might be a voltage regulation problem, any suggestions? I also noticed it occurrs when the ram and CPU are both being stressed. however i have yet to test the video cards.

Also during these tests no componant reach a temp higher than 38c. highest being during memtest86 which the ram did hit 37.6c. Temp is more than ok

 

[omnage]

Noticed something odd during the test that made me concerned. RAM clock speed was set to auto in the bios and windows posted it as 2000mhz. My ram is only rated for 1600. I changed the setting to match manufactures speed and have since fixed the issue. I can also under rate the ram at 1033 but no reason to do so. could that be all it was or is that a temporary fix? it leaves me puzzled that the bios did not auto detect the issue.

 

[panais]

Noticed something odd during the test that made me concerned. RAM clock speed was set to auto in the bios and windows posted it as 2000mhz. My ram is only rated for 1600. I changed the setting to match manufactures speed and have since fixed the issue. I can also under rate the ram at 1033 but no reason to do so. could that be all it was or is that a temporary fix? it leaves me puzzled that the bios did not auto detect the issue.

Yes,setting manually the right clock speed in some cases is the solution.