6970 Black Screen or Vertical Lines

Hello all.
After being almost rudely ignored on two other forums about graphics cards, I decided to come back to you guys, even though you're not a website that specializes in graphics cards. Before I start, just to note - I've heard the problem I have can be caused by a hardware issue and not only Windows 7.

SO HERE'S THE DEAL:
I changed my good ol' perfectly functioning 4850 for a 6970. The first issues cam when my computer started shutting off randomly, I have now taken care of this problem and upgraded my 500W Stock Supply to a CM 750W Bronze. Unfortunately, there is another issue which is driving me so mad, I'm ready to through my entire PC setup out the window. The card is giving me the vertical lines problem. This is a well known problem for some 4850's (even though my old 4850 had no such problem) and for almost all 5000 series users. The problem is rumored to be caused either by a bad memory clock, voltage or fan speed setting, but I've spent about two hours messing around with those (both overclocking and underclocking) and have only gotten things worse - the vertical lines seem to be gone, but now my computer crashes with the much more unpleasant BLACK SCREEN. You game and then - BLACK SCREEN! What I've noticed is that the audio sometimes continues normally, but Alt+F4-ing and Ctrl+Alt+Del-ing doesn't do anything.

Right now I have reset everything (dirvers, card, OC settings) to default and I have no idea what to do.

Just a note - if I can't gat this fixed in a few days I probably WILL go crazy and ragequit not only PC gaming, but the computer world itself. I just can't help but be baffled by how an estimated of a 550$ PC upgrade causes NOTHING BUT PROBLEMS.

Anyone know what to do?

After doing a quick search on your problem I can feel your pain. Seems their is no shortage of complaints.
Have you tried contacting the manufacturer to see what their take is on this seemingly wide spread problem?
Fabe

Well since you replaced the power supply I would be pointing to graphics, sounds like the card is just deciding to shut off on you randomly.
I would RMA a card doing that myself wouldn't even think twice about it.
Especially a flagship price tag model like the 6970.

There is of course the obvious check you can preform...
Just find another computer you can plug it into and test it on.
After my original 5850 failed on 3 different machines they finally agreed to the rma
The replacement has a newer ver number on it, and has not given me the slightest problem.

This is something to note, especially if you buy it in a physical store. On the label there is a place on the big white factory sticker that indicates the version number.
If it's 1 or 1.1 ... I would generally look for something else.
The 5850 they sent me that finally worked is a ver 4.1

@thefabe - NO WAY! I am NOT contacting the manufacturer or the store from where I got the card. If there will be any fixing I will have to do it myself. Also, right now I literally can't say I have more than 30$ left from all the cash I spent on my PC... At the store, if this turns out to be a serous issue I will also be probably asked to leave the computer for testing for a few days which would suck. No computer for me means no job, no work, no entertainment, no internet, no pr0n, no facebook, no news, no online gaming NO NOTHING! Plus, if I fix this by myself I'll really feel more confident about my Pc skils and all... lol.

@ManOwar2 - Here is the link to the card on newegg. All I see is one guy who seems to be much happier than I am...

AS FOR THE PROBLEM:
I think I am slowly starting to get to the bottom of the problem. Here's a quick GPU-Z log of me playing GTA IV until a Vertical Line crash appeared (Attachment).
GPU-Z Sensor Log.txt

You guys see what I see? Core clock jumping like crazy from 250 to 800 or sometimes even to 920... Same with the GPU memory clock - incredible variety and randomness going from 150 to 1450...

As for the temperature, I've heard that 80 degrees is the max for this card and with an average of 72 (keeping in mind that only GPU temp 3 is at 78 degrees) the card SHOULDN'T BE OVERHEATING! My case is also a tower standard-ish case, but with two cooler master additional fans... Also, if you have any doubts about the card cooling here is what it looks like.

In the end, is there any way I can have a constant Core Clock and Memory Clock setting as is that by any means dangerous?

P.S.: **** my warranty right now. **** it hard. I AM willing to do everything to get this card to work!

The onboard nvidia bios will generally shut the card down before it actually reaches failure heat levels.

Maguscreed said:

The onboard nvidia bios will generally shut the card down before it actually reaches failure heat levels.

Umm... you mean ATi, right?

And as for the high differential in core clock and memory clock? Do you think this my be the cause? (P.S.: The voltage is also quite weird...)

lol I probably should have just said video card bios, because it's true for either one.
Sorry I haven't slept in about 29 hours now starting to get a little bit loopy.

@Maguscreed - I see. No problem. People make mistakes all the time - like I just did by PURCHASING THIS CARD!

Vertical Lines - GTA IV.txt
Did a constant voltage, constant core clock, memory clock (log above proves it). Got a temp a bit above 70 and crashed AGAIN. This time I got the vertical lines instead of black screen and what I have noticed is that when I get vertical lines the audio continues for about 9-10 seconds and then my computer completely freezes. The more stuff I do the more and more I feel I have a faulty card.
Darn... and to think I was actually looking forward to this particular upgrade...

Since you don't want to deal with the warranty, normally I would not recommend this: Try removing the heat sink/fan assembly (be carefull).....clean the chip & the heat sink & any ram it contacts, with 95% rubbing alcohol. Then once all clean: take a good quality thermal paste, (Artic Silver or better) take the heat sink put a good spot of paste on it: take a credit card and work it back and forth till you have a good, even, thin layer all over the contact surface, then put a small amount on any chips the heat sink contacts. (not too much...) Reinstall amd see if that doesn't help those temps.......

This could possibly void the warranty.....