Intermittent Boot problems

atheos · 11 Jan 2009 at 17:37

Hey all, This forum sure is a wealth of knowledge so i have a question / looking for advice on a little problem i have.

I bought a second-hand 8800GTS(G80) before xmas. Everything was fine until new year, when i noticed some artifacting in games. A day later while playing GRID i got a horrid BSOD with the dreaded "NVLDDMKM.dll" error. I rebooted and ran 3dmark 06 and got bad artifacting and when i rebooted again i got no display on monitor.

Some people say in forums that if u blue screen with this error it can damage your motherboard. I tried one (the main xfx) card in both pci-e slots and no signal to monitor but the fan turns on and my mobo boots(FF). I tried the second hand card i bought(club3d) and monitor turned on, all tests ran fine, 3dmark, vantage, furmark etc.

I have RMA'd the (main) card back to the e-tailer i bought it from. It was almost two years old, so i was just in time as it had a 2 year warranty.

Excuse the long story

but the problem is every couple of days (booting from cold and not every day, maybe 1 in 3) when i turn on the pc, i get 1 long beep with three very short beeps, which i am lead to believe is a vga error. i turn the pc off and back on and i get one quick beep and pc boots and all passes all tests. Before i went sli (maybe a coincidence) my pc was rock solid and i had never overclocked

My temps are fine as i have everest displayed on my g15 keyboard.

Do you guys think that NVLDDMKM.dll BSOD could damage your mobo or that it is possible that my motherboard / main pci-e slot could be in some way damaged?

If you dont ask you dont learn,

Thanking you in advance

atheos · 12 Jan 2009 at 16:05

Same thing this morning, turned pc on. first two tries the beeps. third time no problem. Has anyone got any thoughts on this at all?

95thrifles · 12 Jan 2009 at 16:08

No, an error shouldnt cause any hardware damage, the worst it could do really is corrupt a sector on your hard drive. Sounds like you gfx card may not be seated quite right, they can be funny sometimes, take it out and put it back in, see if that fixes it

Admiral Huddy · 12 Jan 2009 at 16:14

If you are experiencing POST beeps then cross reference these with www.bioscentral.com making sure you know which BIOS you have (this will be displayed on the first line of the POST or check the motherboard manual)

If you feel that it is a display problem and it's a coincidence that it's been since SLi, then it stands to reason this could be the cause.

Remove the second card for a week or so and see if the problem re-occurs with just a single.

Swap over the cards in single mode - If one card works but the other doesn't then one card is faulty!

Check the PCI port by switching the single card to the other slot and test. If a known working card works in one but not the other, you have a faulty mobo.

Replace the bridge - if that works with SLi, then it was the bridge.

Check the for latest drivers.

Have you checked the PSU rails for stable voltages?

I might be worth checking for BIOS updates in either case too.

ChrisLX200 · 12 Jan 2009 at 16:21

If I hadn't seen your PSU spec I would have said it was related to that, gives the symptoms of underpower when booting from stone cold (highest power draw for longest time). Still maybe check your power - all plugs to GPUs are secure?

atheos · 12 Jan 2009 at 16:21

Thank you for your reply guys. One of the cards has gone back to be rma'd so i am left with one card and unfortunately dont have another card to try.

My PS is an enermax galaxy 1k so i think its ok (i hope).

I will try running the card in the second pci-e slot later on.

Admiral Huddy · 12 Jan 2009 at 16:24

atheos said:
My PS is an enermax galaxy 1k so i think its ok (i hope).

Then test it.. This may help.

Diagnosis is process of elimination.

If you are into building PCs, then it's always worth having cheap spares floating around, just in case. Even if it's a bog standard £15 PCI card, for example.

atheos · 12 Jan 2009 at 16:31

Good point Admiral Huddy

. I will try to arrange a multi meter from someone or buy one soon. Time to order a dirt cheap card i think as well for troubleshooting purposes

atheos · 14 Jan 2009 at 21:56

Just to complete the post, i swapped my card into the second pci-e 16x slot and no problems at all. Just have to make do with a single card until my next build :mad:

The previous night while watching some tv shows on pc, the screen lost signal 3 times and the 4th time i got a screen lock up and my temps on my g15 for the card was 0.

So to sum it up, i think that the pci-e slot nearest to the cpu is faulty. Strange how it went the same time as my card eh! (got a 9800gt as replacement so all not lost)

Puters eh!