• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Fried i7 920?

T

TigerLord

[H]ard|Gawd
Joined
Mar 11, 2007
Messages
1,085
My full specs are in my signature.

This morning I received a reference Asus 6950HD card. I proceeded to flash it with a 6970 Bios with success. I noticed the first time I boot up there was a small spark at the PCI-E connectors level. Everything worked fine for over six hours, and I began to blue screen. The first message had to do with Kernels, but then it would just say "A process or thread crucial to system operation has unexpectedly exited or been terminated."

So I flashed the card back to the original 6950 bios I had backed up this morning, still got the same error. I figured it couldn't be... I swapped the card with an old 8600GTS I keep as backup. Same symptoms: after booting into Windows, 1-2 min later on idle or if I try to do anything, I blue screen and get the message "A process or thread crucial to system operation has unexpectedly exited or been terminated." So I booted back in BIOS, set everything back to default (stock speeds, voltage, etc) and tried it again. Exact same problem.

Then suddenly, I began playing urevideo (720p mkv)I bought my 920 used here in December, but the original owner said he had never OC'd the processor, and it's been working fine until today.

I have no idea how to troubleshoot this. This message is consistent with failing overclock, but why would I still get it at stock settings? Could I have inadvertently fried it, somehow? How can you check that?

I'm not sure what to do to troubleshoot stuff at this point.

Please help!
 
probably fried the PCI-E slot on the motherboard. since you indicated sparks.
 
86 5.0L said:
100% = dead,
0% = no problems

you lie somewhere inbetween. Try another slot on the motherboard.
Click to expand...

I think this is the problem you have, but you could have also messed up something on the motherboard. I doubt you hurt the CPU with that. Good luck!
 
Been getting 0x00000074 startup BSOD BAD_SYSTEM_CONFIG_INFO blue screen errors, and whenever I reboot after a crash, I'll get a System Disk Failure... I think the entire mobo might be fucked.

I'll try putting the card in another PCI-E slot and see what happens.

I don't understand, either. The 6970 BIOS I used was to unlock the shaders only, why would it fry the RAM and subsequently, the mobo? Crazy!
 
TigerLord said:
Been getting 0x00000074 startup BSOD BAD_SYSTEM_CONFIG_INFO blue screen errors, and whenever I reboot after a crash, I'll get a System Disk Failure... I think the entire mobo might be fucked.

I'll try putting the card in another PCI-E slot and see what happens.

I don't understand, either. The 6970 BIOS I used was to unlock the shaders only, why would it fry the RAM and subsequently, the mobo? Crazy!
Click to expand...

Please download and install this
http://www.nirsoft.net/utils/bluescreenview_setup.exe

The files in red post their names here
 
I disconnected all other drives because every time I rebooted from a crash, the BIOS would change boot order. So I'd either get a DISK SYSTEM FAILURE or it'd ask to boot from the AHCI CD-ROM. 25% of the time it will say my RAID ARRAY is not healthy.

I just got a new blue screen, KERNEL_DATA_INPAGE_ERROR.

I am now almost sure my mobo is the issue. I will go buy the cheapest 1366 board I can find tomorrow and see what happens. Is there any chance this is SSD related?

Will try to install that program if I can manage to stay in Windows long enough.
 
atikmdag.sys atikmdag.sys+90817 fffff880`05a2b000 fffff880`0623c000 0x00811000 0x4cef1f94 11/25/2010 10:46:44 PM
ntoskrnl.exe ntoskrnl.exe+6fca9 fffff800`03468000 fffff800`03a45000 0x005dd000 0x4cc791bd 10/26/2010 10:43:09 PM Microsoft® Windows® Operating System NT Kernel & System 6.1.7600.16695 (win7_gdr.101026-1503) Microsoft Corporation C:\Windows\system32\ntoskrnl.exe

Red items.
 
Barely had time to post this and it crashed again... posting from laptop still.
 
I bluescreen in safe mode too... I thought GPU drivers were'nt loaded in safe mode?
 
System is clean, as it should as I run Avast 5.0 Pro, AntiSuper Spyware and MalwareByte.

It's the mobo, only thing left. I tried installing windows on a brand new HDD, blue screen during install.
 
Might want to reflash the bios if you can just to be safe. Otherwise yea you probably fried something on the mobo. I have a dead one sitting here next to me.
Older gigabyte for the old quad core.
 
This is batshit crazy... I remembred the Classified3 has a RESET CMOS button on the mobo's PCB. I reset it, then tried a fresh install on a new drive. Whereas I'd get constant lock-ups 2% in, I'm now at 98% without a single freeze or slowdown.

Could resetting CMOS really have solved this? A shortcircuit messing up the bios? Is it even possible?
 
Definitely interesting if it does indeed solve things. Its a miracle and you are lucky if there are no further problems.

I would still be cautious and be prepared to buy another motherboard or RMA your current one just in case.
 
TigerLord said:
System is clean, as it should as I run Avast 5.0 Pro, AntiSuper Spyware and MalwareByte.

It's the mobo, only thing left. I tried installing windows on a brand new HDD, blue screen during install.
Click to expand...

Ok lets do the following
Go to start, enter msconfig, right click properties->run as admin

Then first tab tick bootlog.
Services tab, tick hide all ms services. Then tick disable all. Then ticks only the ones related to your gpu.
Then startup services Untick all, apply reboot. test it
 
Stop messing with the board and RMA it.
Before the magic smoke takes out your CPU, hard drive and power supply.
Playing with a board that had an electrical failure is VERY stupid.
 
86 5.0L said:
100% = dead,
0% = no problems

you lie somewhere inbetween. Try another slot on the motherboard.
Click to expand...

he, sounds like you're around the 95% mark.

and yes, stop booting the motherboard, if theres a short somewhere you're only make it worse with every powerup
 
Yes, I stopped. Figured I should give it a try before entering the whole RMA fiasco.

My biggest concern is actually the RAID array. I plan to buy a Sabertooth X58 while I wait for the Classified3 replacement, because EVGA doesnt do cross-ships to Canada anymore. So it'll take 4 days to get to them, another 2-3 days for them to ship it out, and another 3 days to reach me. 10 business days = 2 full weeks. Can't be out of action this long.

Is the RAID array specific to the controller? How could I back it up now?
 
I picked up a MSI X58M to serve as an intermediary between RMA.

It has the same ICH10R chipset as my Classified3. Can I just plug my SSD in and it'll recognize the array?
 
Simple answer with a RAID you need to make sure that you have the same controller or configuration to carry over a RAID set up from one system to a replacement, unless it is a RAID 1 (mirror).
 
I changed the new mobo in, but it seems the RAID array might have been damaged. One of the SSD drive comes up as "Error Occured".

It is repairable?
 
In a shocking plot twist, I am getting the same "A process or thread crucial to system operation has unexpectedly exited or been terminated." blue screen with the brand new mobo.

This leaves a bad PSU or bad CPU.
 
I tried a new mobo, blue screens.
I tried a new PSU, blue screens. But the NEW PSU somehow allowed me to backup files and open programs (like unregistering Adobe suites, etc).

The RAID controller kept reporting one of my OCZ Vertex2 as "Error Occured(0)". After backing my file I rebuilt the RAID array, and am now trying to install Windows 7.

Is it possible the SSD drives are at the source of all my problems?
 
Spoony, thanks so much for your help with everything. It wasn't what you suggested, but I appreciate you tried!

The problem laid in the SSD drives. I looked further into it and I *did* flash the 6950 fully, not just unlock the shaders. I did not know the RAM timings and voltages were different. There is now no doubt in my mind that this was the cause of the very first BSOD. After a few blue screens and finally flashing the original 6950 BIOS back, the GPU was fine, but the SSD drives were corrupted. The OCZ forums are FULL of people having controller issues or bad drives, or corrupted drives that require special steps. The SSD I used in my laptop suffered the time warp problem, which was very frustrating, and require some further steps.

I should have picked up that when the RAID controller reported "Error Occured(0)" in status for one of my SSD drives, something was wrong. But I never thought SSD drives could have been responsible for so many blue screens. I've only began to use SSD since December. It's clearly a technolgy that is still very young and presents many problems. Worth the hassle imho, and this whole experience will have taught me A LOT.

So, what did I do? I had to "fix" the SSD first, then rebuild the array, then make a fresh install. Since then I have popped everything original component (except the 6950 which I sent back to get a 6970 instead). The mobo is fine - there will be no need to RMA it. I am relieved and happy, but still pissed that I bought a 200$ mobo for nothing. Live and learn, I guess!

Here is what I did, step by step, in case anyone ever wonders here.

1. Made a fresh install of Windows 7 on a normal, non-SSD hard drive (BIOS SATA in AHCI mode).
2. Downloaded OCZToolbox in preparation for the next steps
3. Shutdown PC, unplugged main hard drive.
4. Popped both SSD drives back in
5. Set SATA mode to RAID in BIOS
6. Entered RAID configuration tool during boot (CTRL+I)
7. Deleted the RAID array
8. Went back into BIOS, set the SATA mode back to AHCI. This is important since I was planning to flash the firmware of the drives, and OCZtoolbox's current version prefers when the drives are in AHCI mode. Also, you cannot flash firmware on SSD drives that have the OS on them, only if they're slaves. A new version is coming out next week that should allow firmware flashing even for OS drives.
9. Shutdown PC, plug back in the original hard drive where you installed Win7
10. Go back to BIOS and make sure first boot drive is the hard drive (not the SSD drives)
11. Boot to Windows with the SSD as slave drives (not in RAID). Once in Windows, I used OCZToolbox to update the FW from 1.24 to 1.32 since I was overdue, then Secure Erased both SSD drives. This is the magical step that "fixes" corrupted drives. A hard reset if you will.
12. Shutdown PC, removed hard drive with Win7 on it.
13. Plugged in my bootable Win7 USB installation key (though you can install from a CD or DVD, OCZ insists that it is much better to install windows from a USB key. Something about SATA controllers and corrupted data... if you dont know how to create a bootable key, simply insert your installation DVD in your drive and copy the content of the DVD onto the key (select all+drag in windows explorer works fine). Then simply plug in your key and set it as first boot drive. Piece of cake.
14. Powered up, went to bios, set sata mode to RAID so I could rebuild the array now that the drives are "fixed"
15. Rebuilt the RAID array
16. Continue booting into Win7 installation.
17. When installation comes up, choose custom
18. You'll see one array of 223GB, click on it. DO NOT MANUALLY CREATE ANY PARTITION!!!! This is very, very important. It fucks everything up.
19. Click Next, that's it. Nothing else. Select unallocated partition then Next.
20. Installation begins. Partitions will be created automtically. When it counts down to reboot when the first step is over, remove the USB key.
21. It will reboot and complete the installation on the SSD
22. Complete the final steps (configuring username, timezone, etc.)
23. Voila, you're done. You've also aged 10 years and feel like you need many drinks.

It is VITAL that ONLY your SSD drives be plugged in when installing Windows 7 on them. Unplug every other hard drive and even DVD/CD drives. SSD are sensititve to sata controllers acting up, so leaving a single option will eliminate any potential issues.

Everything has been fixed and resolve. I spent over 300$ to fix a mistake that was made because I thought I could save 100$ by getting a 6950 and unlock it instead of 6970. I don't get it, though... techPowerUp is a reputable site. Why isn't there a huge, big fucking warning in the article that flashing the bios can fuck up your card??? The shader unlocking method isn't even mentioned unless you dig through the 100 pages thread on their forums. I take full responsibility for what I did, but this is still apalling. Like I said, live, and LEARN.

I'm just happy I found the culprit and everything is fixed. Now I'm getting my U3011 and XFX 6970 on Monday, and will consider this experience a LESSON LEARNED.

Thanks to [H] for helping me out as well!
 
Last edited:
12+hours stability testing later and all is swell.

Upgrading the SSD firmware also gave me a drastic 23% performance increase on writing across the boards, so I guess something positive came out of this!
 
TigerLord said:
Why isn't there a huge, big fucking warning in the article that flashing the bios can fuck up your card??? The shader unlocking method isn't even mentioned unless you dig through the 100 pages thread on their forums. I take full responsibility for what I did, but this is still apalling. Like I said, live, and LEARN.
Click to expand...

pretty sure its common knowledge and is a warning in all the flashing software that I've used.

http://hardforum.com/showthread.php?t=1572460

A Word of Warning: Do not attempt this unless you can fathom the possibility of a damaged GPU as the end result. Some people that have flashed have had their memory hosed by doing this flashing method. Even after flashing back to the original BIOS they still experienced artifacts and were unable to flash into pre-artifact condition. Although one can use the fail-safe BIOS as a means to flash the primary back to its stock form, this modification technically voids one's warranty.
Click to expand...
 
It is common knowledge of course, but if the risk is big enough that sparks will fly out of your card, it requires additional warning, imho. For instance, I know overclocking can potentially damage my hardware, but if you stay within safe limits, you have a 1% chance of fucking up.

I just flashed a stock bios and I got sparks... we're not on the same level of danger here anymore.
 
TigerLord said:
It is common knowledge of course, but if the risk is big enough that sparks will fly out of your card, it requires additional warning, imho. For instance, I know overclocking can potentially damage my hardware, but if you stay within safe limits, you have a 1% chance of fucking up.

I just flashed a stock bios and I got sparks... we're not on the same level of danger here anymore.
Click to expand...

anytime you mess within your PC theres a chance to fuck things up, you rolled the dice and lost, ESD or PSU discharge got the better of you. you hit that 1% live and learn :)
 
You must log in or register to reply here.
Back
Top