• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Really fed up now....

iB CoMpL3XiTy

iB CoMpL3XiTy

iB CoMpL3XiTy

Associate
Joined
10 Sep 2010
Posts
1,533
Location
Cornwall, England
Hello fellow enthusiasts, I'm having a crap time atm

Please read this http://forums.overclockers.co.uk/showthread.php?t=18252960&page=3 to get the background info

The reason I'm not carrying on with that thread is I don't know the forum's rules on reviving an old thread.

To cut a very long story short, I was getting Bluescreens caused by nvlddmkm.sys (or whatever it is) so I thought that by re-installing windows and installing two brand new XFX HD6950's would solve the issue. Nope.

Now getting bluescreens caused by atikmdag.sys

I've replaced every single component in my PC since the issue first arose and installed windows 7 coutless times and from two different CD's. All drivers and BIOS's are up to date.

my patience is wearing thin and so is my wallet, any help is much appreciated, just like to say thanks to the people that took the time to try and help me in my previous thread.


Thoughts and opinions?
 
Timeout detection and recovery error? Yes, been there. Massively, massively frustrating.

Everything needs to be run at stock. No overclocks allowed on cpu, gpu, or ram till you find the fault.
PSU 12V rail needs enough juice for the cards your using (and a bit more to be safe).
Ideally like the lad above said, run one card on its own.
Memory needs to be tried in each individual slot (one stick on its own at a time).

Consider than sometimes TDR's can be caused by incompatible components (not just faulty ones) and if you have some brand new kit mixed with some really old kit in the same machine, that can cause it too.

Also you can:

Try the card(s) in someone elses machine
Try memory in someone elses machine.
Run the Power supply on a tester to verify outputs.

I know where your coming from mate, I spent over 500 fault finding, turned out it was my graphics card that was faulty in the end.
 
Ive read about disabling ulps in the registry,failing that i would reinstall windows and nothing else not even updates and use the drivers from the disc that came with the card, then something to test the system with
 
@muggs, currently testing with just one HD6950 atm, just played abit of DoW2R (What a disappointing game, orginal was much better)

I noticed when I took the first card out that one of the GPU screws on the second card was really loose, needless to say I've tightened it now, bit concerning though. Both cards are brand new.

@Ladiesman, Sorry to read that man :( Watch some Lee evans

@Michaeljcox, Everything is at stock, has been for about 3 months, sad times.
Using a NZXT Hale90 1000W PSU, also tried a Corsair HX850, both should be plenty.
Memtested the RAM for 9 hours, also tried different RAM that I know is working perfectly as my brother is currently using it.
Everything in my PC is new, except the motherboard (although I did try a new EVGA SLI 3 board aswell) its an X58 build you can find the specs in the OP of the thread I linked to.
I don't have a PSU tester but my BIOS is reporting voltages well within safe limits.

I would hate to hazard a guess at what I've spent trying to get a working PC together.

@old gamer, I've had the issue whilst installing Windows 7 and also whilst posting, can't see it being a software issue, could be wrong though...

@spleenharvester, eh? whaaa? :P
 
Do you have a new HDD?

I think my old HDD is starting to introduce random errors - it's caused BSODs on multiple Windows installs with my old IP35/Q6600 combo and my P8P67/i5-2500k combo - stuck another drive in there, it's been fine for the last few days.

However - the old HDD doesn't fail any error checks!!
 
iB CoMpL3XiTy said:
@spleenharvester, eh? whaaa? :P
Click to expand...

Start menu, Right click "Computer" then properties, advanced system settings, system protection, create restore point.

To run verifier, open a command window, type "verifier" on its own without quotes, select "Create custom settings" (for code developers), Select individual settings from a full list, select all. Then "Automatically select all drivers on this computer".

Will make it slow as **** and could need a few reboots before it catches something, but if it truly is something driver related it'll catch it :) It'll sit on the loading screen for a while so don't be unnerved.

I had problems with TDR_TIMEOUT_ERRORs and random BSoDs, turned out to be PxHlpa64.sys causing I/O errors. Bizarre stuff.
 
Confused said:
Do you have a new HDD?

I think my old HDD is starting to introduce random errors - it's caused BSODs on multiple Windows installs with my old IP35/Q6600 combo and my P8P67/i5-2500k combo - stuck another drive in there, it's been fine for the last few days.

However - the old HDD doesn't fail any error checks!!
Click to expand...

Yeah, had a 500GB sammy spinpoint F3 die on me, replaced it with a 80GB OCZ Vertex 2 and a 1TB sammy F3
 
Had a similar issue with my old system, reinstalled windows many times on several drives and replaced everthing bit by bit.
It seemed the last bit I replaced was causing the issue :) The motheboard !
 
Try this:

1: Disable the Windows Advanced Settings which cause a restart automatically if any error is found. Waste of time unless your a developer most errors it can recover from as the drivers allow for this anyway. Sometimes it can reboot for minor errors. Many serious driver errors are recoverable by the drivers they auto reboot without causing Windows to reboot unless you have that restart automatically setting enabled.

2: Run your Ram below spec so its much lower than the rating. Core i7 on some mobo/ram combinations are very sensitive to Ram timings & Bios settings.

3: Do not bother overclocking anything. Its a waste of time/money nowadays. You gain a few % at the expense of hardware failure/damage. Really ask yourself what real world difference do a few % make!!! Doubt you will notice (but your wallet would if it goes wrong!). Overclocking is an art & most serious ocers use water or LN2 on everything but even then they are careful & can still damage components.

4: Build the PC using 1 stick of Ram only. If possible test all Ram for errors. Just because it works on someone else PC means nothing they may have different Bios settings than you. Its possible for a faulty Ram or PCI-X slot to cause BSOD's. Try to use the minimum of hardware only. 1 gfx card etc etc then add after it becomes stable for a few hours under load from a looped benchmark run. Try to isolate components so you can see what if anything is faulty.

5: Flash the motherboard Bios to the default stock version. Do not change any settings whatsoever.

Logically if you say you have replaced everything then try to get someone else to build this for you as the chances of 2 different sets of hardware having issues are tiny unless you have done something consistently wrong.

How long have you been building PC's????
 
Last edited:
Just to rule out some arcane stuff. Have you tried connecting PC to a different power socket outside of its normal room, or the complete system (including your monitor, mouse, keyboard) at a friends?

Thoughts I had, but probably have no use as not a technical support guy.
  • Check the PSU OEM, e.g. are they both rebranded Seasonic models. Maybe try a different OEM with multiple rails.
  • Check Corsair Dominator revision number and identify what memory chips it uses by that. May be same chips as previous memory as Corsair just source what is available during time of manufacturer, and supplies change during the product lifespan.
 
spleenharvester said:
Start menu, Right click "Computer" then properties, advanced system settings, system protection, create restore point.

To run verifier, open a command window, type "verifier" on its own without quotes, select "Create custom settings" (for code developers), Select individual settings from a full list, select all. Then "Automatically select all drivers on this computer".

Will make it slow as **** and could need a few reboots before it catches something, but if it truly is something driver related it'll catch it :) It'll sit on the loading screen for a while so don't be unnerved.

I had problems with TDR_TIMEOUT_ERRORs and random BSoDs, turned out to be PxHlpa64.sys causing I/O errors. Bizarre stuff.
Click to expand...

Thanks for explaining it :)

I did try this and Windows took about 5 minutes to load and then about 2-3 minutes after it loaded the screen would go completely black, but remained on, I left it like this for over an hour and decided to restore from a backup.
 
AWPC said:
Try this:

1: Disable the Windows Advanced Settings which cause a restart automatically if any error is found. Waste of time unless your a developer most errors it can recover from as the drivers allow for this anyway. Sometimes it can reboot for minor errors. Many serious driver errors are recoverable by the drivers they auto reboot without causing Windows to reboot unless you have that restart automatically setting enabled.

2: Run your Ram below spec so its much lower than the rating. Core i7 on some mobo/ram combinations are very sensitive to Ram timings & Bios settings.

3: Do not bother overclocking anything. Its a waste of time/money nowadays. You gain a few % at the expense of hardware failure/damage. Really ask yourself what real world difference do a few % make!!! Doubt you will notice (but your wallet would if it goes wrong!). Overclocking is an art & most serious ocers use water or LN2 on everything but even then they are careful & can still damage components.

4: Build the PC using 1 stick of Ram only. If possible test all Ram for errors. Just because it works on someone else PC means nothing they may have different Bios settings than you. Its possible for a faulty Ram or PCI-X slot to cause BSOD's. Try to use the minimum of hardware only. 1 gfx card etc etc then add after it becomes stable for a few hours under load from a looped benchmark run. Try to isolate components so you can see what if anything is faulty.

5: Flash the motherboard Bios to the default stock version. Do not change any settings whatsoever.

Logically if you say you have replaced everything then try to get someone else to build this for you as the chances of 2 different sets of hardware having issues are tiny unless you have done something consistently wrong.

How long have you been building PC's????
Click to expand...

I've disabled the restart thing, doesn't make any difference at all.

I've tried using my RAM at 1066Mhz AUTO timings and Voltage, aswell as its rated values, done this with both sets of RAM.

I don't agree with your third point, however I'm not going to get into a debate about it. Besides which like I mentioned in the OP everything has been at stock for 3 months.

I don't have anything to put a BIOS onto atm, my Brother has my USB stick, I have tried 3 different motherboard BIOS's though.

Been building PC's for over 6 years now.
 
iB CoMpL3XiTy said:
I've disabled the restart thing, doesn't make any difference at all. That eliminates software/driver issues then.

I've tried using my RAM at 1066Mhz AUTO timings and Voltage, aswell as its rated values, done this with both sets of RAM. Try using very low slack timings as in way below it rating. My Core i7-965 would not work properly with default or auto timings not all mobos work properly with Auto Timings in certain CPU's due to Bios bugs as they are never ever tested with all Ram timings/brands.

I don't agree with your third point, however I'm not going to get into a debate about it. Besides which like I mentioned in the OP everything has been at stock for 3 months. If something is damaged that will make no difference the damage is done certain high amp loadings will cause a crash regardless.

I don't have anything to put a BIOS onto atm, my Brother has my USB stick, I have tried 3 different motherboard BIOS's though. Buy a USB stick or use the Windows Utility to flash it as most Mobos have dual bios nowadays anyway. I had a corrupt bios on an old DFI mobo & only 1 bios was corrupt so the mobo had an intermittent issue where it would find the corrupt bios & crash then reboot & work until it found the corrupt bios on boot again etc etc
You would be amazed how easily Bios corrupt especially when overclocking.


What Motherboard do you have anyway??

Been building PC's for over 6 years now.
Click to expand...

Look at this logically if its not hardware & its not software it only leaves a corrupt Motherboard Bios!

When/if you flash your bios back to default make sure the first thing you do after powering back on is load the bios defaults from within the Bios as this prevents nasty Bios bugs you cannot even see from sticking Bios values to something different to the Bios screen. In my 19 years experience this happens a lot!!
 
Just tried with my RAM timings at 10-10-10-30 800Mhz 2T 1.64V Same issue still occuring
Not overclocked my 2 week old Core i7 960 or either of my HD6950's, no overclocks on any of the replacement components.

I don't like the sound of flashing my BIOS in Windows, I'll have to buy another USB stick or get mine from my brother.

I'm using a Gigabye GA-EX58-UD5

I'll be removing the watercooling for the CPU tonight and replacing it with the intel stock cooler.
 
iB CoMpL3XiTy said:
Just tried with my RAM timings at 10-10-10-30 800Mhz 2T 1.64V Same issue still occuring
Not overclocked my 2 week old Core i7 960 or either of my HD6950's, no overclocks on any of the replacement components.

I don't like the sound of flashing my BIOS in Windows, I'll have to buy another USB stick or get mine from my brother.

I'm using a Gigabye GA-EX58-UD5

I'll be removing the watercooling for the CPU tonight and replacing it with the intel stock cooler.
Click to expand...
I have the exact same mobo with my Core i7-965 PC (bought it Nov 2008)!!!

I reckon you have the same issue I had with Ram timings as that mobo really does not like certain timings at all & the Dual Bios can easily corrupt. I flashed mine using the Gigabyte Util it works fine & even if it failed you hit a function key & it will ask for the original CD which came with the mobo on power-on. Look up how you do it in your manual I cannot remember the key its hold down F12 perhaps after power on.

Follow this guide exactly on their official Bios Forums.

http://forums.tweaktown.com/gigabyte/27576-bios-flashing-how-qflash-guide.html

Its very important to ensure that the "Keep DMI Data Disabled" box is checked as thats what usually causes a corrupt Dual Bios to be hard to fix as without that little step all your doing in a loop is copying the corrupt Bios over both Dual Bios slots.

I think that may well be your issue as I had a lot of trouble in Nov 2008 with my then brand new Core i7-965XE + that mobo caused by Ram timings & a corrupt Dual Bios.

This thread has links to all the various UD5 bioses & its official as Gigabyte pay tech staff to monitor it but chose to let tweaktown host it !!

http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-beta-bios.html

So to recap:

1: Flash your Bios back to factory (you can do this from the CD). Uncheck keep DMI data.

Make sure you also short the Mobo Pins for clearing the CMOS data & also remove the battery for 30 secs or so as this prevents Bios corruption from being retained. Follow the procedure in the manual for this they have all the steps written out follow them to the letter even if it seems crazy just do it!

2: After flashing & power down. Power on & go straight back into the Bios.

3: Now load the default Bios settings then in the Bios menu save the settings.

4: Now exit the Bios then power off. Wait 30 secs or so & power back ongoing straight back into the Bios again.

5: Save the Bios again in the Bios menu without changing anything.

6: Power down. Wait 30 secs power on & it should have the default bios on both Dual Bios slots.

This will definately give you a fresh bios as per factory settings now with confidence you can see which if any components are faulty but I reckon this will fix it as those particular mobos are very easy to corrupt as they are the earliest X58 based mobo so I think Gigabyte did not do enough QA on the Bios as they rushed them out. Mine has been 100% stable for almost 29 months now after I found out the issue was the mobo is not resilient enough even with Dual bioses!!
 
You must log in or register to reply here.
Back
Top Bottom