Sapphire X800Pro Direct Draw errors? HELP

I have a quandry. I have a failry new Sapphire X800Pro, never modded overclocked (beyond running the ATI Tool test), or flashed. About a week after placing the card in my Asus P4C800E-dlx/Prescott 3.0E rig, things went haywire.

Whenever I'm running DX9 benchmarks, and especially Futuremark benchmarks, the screen will simply crash, reverting to 640x480 horizontal/vertical line-madness. As if it simply lost rendering control. When I checked to see if the card was seated properly, moving it a tad had an effect on the screen, it didn't clear it up, but it kind of "re-set" it. Then I re-booted and it was fine. I tried another video card in the AGP slot to see if the AGP slot may have had a short, but it didn't. This ONLY seems to happen when I'm running Futuremark, but now it happens running the new Gun Metal 2 DX9 benchmark, as if the problem is worsening?

Again moving the card a bit will re-set it and re-booting brings the video back, but I don't understand how a "short", would only be triggered by 3DMark benchmarks? By the way, the temp of the card hasn't risen above 35C as there's a larger fan blowing in the open case, and the room is AC cooled.

I've tried re-formatting, re-installing, using different drivers 4.6 and 4.7, and am at my wits end. I've checked the card for any obvious shorts, the external power molex is fine. The card is securely seated in the slot, and I've tried re-seating it several times. Before I send the card back to ATI which I really hate to do, because the ATI tool revealed this card to be the best overclocker I'd ever seen, (although I haven't ran it overclocked just yet). Anyone out there know what's happening? The attachements are photo's of my screen when the anomaly occurs.

Thank you

Is the rest of your system overclocked?

How far have you raised fsb?

I believe the nf3 -250 gb have a agp/pci lock so that's not the issue...

Could be a bad card...

That looks like my BIOS screen when using my fried 8500

Try running at the lowest AGP speed possible (like 4x instead of 8x), and see if the problem clears up. Also, make sure your AGP bus is running at 66MHz, and no more (easiest way is to run your system at stock if it is currently overclocked).

Does this only happen in 3DMark and GunMetal? What other games/benchmarks have you tested with?
JigPu

Have you tried using DriverCleaner then reinstalling your Vid drivers.

Sorry I should have been more clear.

The current system this card is in is as follows;
Socket-478 Prescott 3.0E
Asus P4C800Edlx (BIOS 1016)
Sapphire X800Pro (at default never really overclocked) Catalyst 4.7
Corsair Twin-X1024 3200XLPRO or
Kingston KHX3200K2/2G Hyper-X3200
Maxtor Diamond Max Plus 9 (SATA150 120GB)
PCPower&Cooling Turbo Cool 510 Deluxe
WinXP SP1

This happens whether the system is overclocked or running at 200FSB (default). It didn't seem to happoen in the Chaintech ZNF3-250 but I hjad to switch to this rig for a article I'm writing. What's bothering me, is that shifting the card "re-sets" it (after rebooting). But what confounds me, is that the anomaly only occurs during these specific benchnmarks? Trying yopur suggestions BRB. Thank you ALL!

What card was in that system before you put in the X800? If it was an Nvidia card I would def say you got a driver problem. The reseting of the screen when you move the card has got me boggled though?

That card is DED! My 9800XT did that also. I tried it in multiple computers with always the same result. Time to RMA it.

Here's the boggle, it doesn't happen in the A64/ZNF3-250 system? driving me mad. I can understand a short, and have no problem RMA-ing it, it's only that when I tested it on the ATI Tool is had the abaility to overclock beter then 100MHz on the Core, and more onb the memory. Although I didn't overclock it, I'd hate to return a card with that much headroom for when I decide to open up the othger 4-pipelines.

I tried running Far Cry, and it crashed again, and checking the temp, it had risen to 48C, and that was afetr the reboot, so it's looka as though it's overheating? I've removed the heatsink, and decided to try thwoing a waterblock on the VPU, to see if this is the problem. Although the thermal paste on the VPU seemed to be evenly distrubted, as if it was making good contact, maybe the fans not running all the time. I forgot to check that though. I did nick my finger when checking it the other day, so I know it's working sometimes.

Weird thing is, the card has absolutely no ramsinks? When I first saw it, it looked like the heatsink covered the memory, but it doesn't.

Does this sound like a heat related issue to anyone, and if so, what do you think could have been the cause? There's fan blowing in my open case (no side panels) in an AC cooled room, so it'ws not heat from the case?