BFG 6800GTOC SLI / DFI LanParty nF4 SLI-D Issues

[Nibiki]

This will be my first post to this forum, so greetings everybody.

The issues I'm experiencing lead me to believe I am dealing with hardware problems, which is why I'm posting here. I've been dealing with a gamut of issues since putting the PC together, and for all my effort I haven't been able to pin down the cause(s). This is going to end up being a long post in an attempt to give you all as much info as I have. First my build:

MOBO: DFI LanParty UT nF4 SLI-D (bios ver 6.00)
CPU: AMD Athlon 64 Processor 3500+ Socket 939
GPU: BFG GeForce 6800CTOC 2x in SLI mode Driver: 7.8.0.1 (currently)
RAM: Kingston DDR-SDRAM 1024MBytes (2x 512Mb) (Frequency 200.9MHz)
PSU: SilverStone Zeus ST65ZF 650W

Now the issues. This rig was built around May/June, and even back then I would have issues with the computer restarting during gaming sessions. I attributed it to heat, so the side of my case went bye bye and I placed a giant floor fan blowing in it. This seemed to fix the lockups/shutdowns, for the most part, but they would still happen. (Even when the fan /wasn't/ blowing on the rig I only ever noticed either GPU's going any higher then 83 degrees C) So I started to shy away from it really being a heat problem, but keep the fan there in case. Of course, I updated games, updated drivers, updated Bios and all that, and the attempts to fix it never lasted long. With the update of my video drivers to (if I recall correctly) 7.6 I got a nice warning that there was insufficient power going to my video card, and that the performance would be turned down to accommodate. I inspected the connections, the PSU, found nothing wrong, and yet no matter what I did with this new driver it would give me the insufficient power notices after playing games for a while. I rolled back my driver, notifications went away..

What happens next marks where I start to get totally lost as to what the cause is. After returning from eating while hosting our LAN, I go to turn my PC on and find the LED lights on my main video card are not on, the fan isn't turning and the card just isn't turning on at all. And this is after a /proper/ shut-down, and maybe an hour away while eating. I'm boggled, but I can not fix it. I get an RMA after the LAN and get a replacement card (of course, not as fancy looking with lights or copper heatsink plates) but at least it works, and I can finally see if it was just that card being faulty that was giving me all the problems. No!

My PC is still locking up after gaming for a while, mostly playing WoW, CS:S and BF2 as I have been. Somewhere within the last month my /chipset/ fan started making noise, off alignment I suppose, and it only got worse and worse. And being that both video cards of course get hot, and both of them practically cover up that chipset fan, I began to wonder if it wasn't the chipset overheating all this time.

As you can see, I have got a lot of issues, and no idea where to begin. What I /have/ done is run the system with only 1 video card in (non-sli mode), and I don't seem to get crashes for either card individually. I have also tested my RAM, and both sticks seem to be fine both individually and paired up. The only conclusions I've been able to draw are either; either the vidcards or the mobo really /can't/ do SLI very well, it's overheating problems on either card or the chipset, or the PSU is faulty. The CPU has a nice big Zalman on it so it ain't that.

With the latest Nvidia drivers (81.85) the rig doesn't seem to crash (at least as much) but I'm getting strange issues when alt-tabing out of WoW and going back in, and I /can't/ alt-tab out of BF2. And again, running either game in non-sli mode solves the issues. (Which in this driver version they made it possible to switch from SLI to non w/o restarting, which is nice, but I /want/ SLI to work!)

I'm utterly at a loss as to what to replace or examine any more then I already have. Any help would be greatly appreciated. As of now I am looking at the MOBO as the main problem and am looking into replacing it. I'm looking at the more recognizable Asus brand's A8N-SLI Deluxe (not Premium because I'm not buying that the 'heatpipe' technology really does that good a job of keeping the chipset cool.) Again, any advice is welcome, wether you agree that it's the mobo and have a suggestion on a replacement, or if it's not the mobo at all.

Thanks very much for reading this long post, and thanks ahead of time for any advice.

*PS: I do /no/ overclocking, no tweaking at all other then your average antialiasing/anisotropic settings.

 

[DoomUK]

With the 81.85's are you still getting shutdowns/lockups? I had some issues with my SLI setup using 78.01.

There should be a utility that came with your mobo to monitor your voltages, use it to check your 12v rail (while in SLI mode, of course). If it drops significantly below 11.90 when under load then you could be looking at a PSU issue.

 

[Nibiki]

The only monitoring/tweaking utility that comes with the MOBO package is 'nTune', and it isn't so detailed as to give me the ability to watch voltages. However, upon this lead I looked up just what a 12v rail is, and found a very nifty program called 'Everest Ultimate Edition', that gives me readouts for all kinds of things, the 12v rail included. After installing it it reports that my 12v rail is at 11.46V /currently/, that's w/o running any games at all! I have yet to figure out how to get the nice On Screen Display to remain on screen even when I'm in a game that's full screen, but if I find out how I will run WoW and BF2 and see how much it drops. I'll report back soon.

*edit*

Ok.. Being unable to keep the Everest OSD up with games in full-screen mode (and BF2 being unable to be in a window at least via the normal video options) my test was limited to playing WoW in a window. But I watched the voltage fluctuate between 11.46 and 11.14, with an average at 11.26. Is this indication the problems are the PSU, and is there anything else I should inspect?

 

[DoomUK]

If your 12v rail drops to as low as 11.14, which is pretty low, I would honestly look at a new PSU... Something will not be getting enough juice, which is thus a good reason for random shutdowns/lockups/BSOD's.

 

[Nibiki]

Ok, I'll definitely be looking into getting my hands on at least a temporary replacement PSU to test against the current one. I just saw your newest post moments after writing the results of my tests w/ the 81.85 driver, so I'll leave those results here to for perusal.

In regards to the question of if the PC locks up/shutsdown when using the 81.85 nVidia drivers, not yet. I wouldn't put it past it because I /have/ had whole days even a few days w/o a crash, but then out of nowhere bam. However, with the 81.85 driver in place, there are two notable issues that are constant. The first is when playing WoW in SLI mode, if I alt-tab out of the game, and go back in, I get a flickering tint at the top 1/3 of the screen, that reaches down to 2/3 of the screen on some flickers. When I enable 'Show GPU load balancing' in the SLI multi-GPU options, I find that I /can't/ alt-tab out of WoW at all. Upon pressing alt-tab, the screen freezes, and I get the windows cursor, but I'm stuck moving it around on a stillshot of my game. Alt-tabing again to get back into the game creates the same screen flicker tint mentioned before. Running WoW while in single GPU mode causes no problems.

Testing Battlefield 2 reveals similar issues. I can not properly alt-tab out of the game. I am allowed only to move the normal windows arrowhead around on a still image of my game. However, alt-tabing back /into/ BF2 from this state doesn't mess up the graphics with a tint-shearing flickering effect like in WoW, however on one occasion the whole screen would flash all black for spells of around a second, a few seconds in between. Restarting the game fixed that. And again, running this game in single GPU mode results in no problems from the testing I've done.

Voltages still fluctuated from what I could decern between shutting the games down and starting them up.