Sign in with
Sign up | Sign in
Your question

semi newb question -- my geforce 6800gt may be dying?

Last response: in Graphics & Displays
Share
February 25, 2007 12:20:08 PM

I've been playing Medieval Total War 2 for months with no problems. Just a few days ago I started running into a strange bug. Basically, just after entering a campaign (starting a new one or loading a saved game) the system would basically slow to a grind. The cursor would freeze in place and whatever sound that was currently playing would continue playing only with a huge stutter, progressing at maybe 1/100th speed. Most of the time I would have to manually power down the computer as alt-f4, ctrl-alt-delete, and alt-tab were all unresponsive. Occasionally after 10-30 seconds of frozen stuttering, the game would recover for a few seconds and then freeze again. In these cases I would be able to quit to desktop eventually. Sometimes the computer would be messed up after quitting out, other times it would be fine. Eventually, I started getting some minor graphical glitches along with the freezing and sound stutter.

I tried updating motherboard (nforce 4) and graphics drivers (which were already up to date). I also disabled other programs and ran msconfig to turn off any unnecessary processes. This didn't help.

I then noticed that when the game freezes, the HD activity light on my case would blink slowly and steadily rather than the normal erratic fast blinking. I tried running a checkdisk and this seemed to fix the problem temporarily. I was able to play fine for a day, then the next day it bugged out again. Checkdisk fixed the problem again.

Meanwhile I downloaded the LOTR online stress test beta and played that for maybe 20 minutes with no issues. That was yesterday.

Today I tried to play LOTR again and met the same error I was getting from Medieval only with more graphical glitches. This time, however, I was able to alt tab to the desktop. What I found was quite strange. I fired up task manager to look at the cpu usage patterns and found that it was basically normal, only every 3-5 seconds there would be a brief (.5-1 second) spike of 100% activity. During each spike my screen would go black.

I forcequit LOTR and the problem persisted. I restarted my computer and the problem went away. I tried running checkdisk again (which took way, way longer this time) and that didn't help this time. Furthermore, the last two times I've tried to use the program I've gotten a hard lock with lots of graphical problems and no ability to alt tab or forcequit. I've had to restart the computer manually. Furthermore, the computer has more graphical glitches in the BIOS and windows startup screen and fails to restart the first time. The second restart works fine. I checked to see if this would happen with Medieval too and I got the exact same error.

I can't even get into the games anymore -- the crash happens at the menu screens.

There are a lot of possible things that could be causing this. It could be something to do with the hard drive, but Everest's SMART check revealed no problems. I'm also not having trouble with any programs other than graphics intensive games.

I'm strongly suspecting the video card here -- if somehow it was leaving bad data behind when the bug happens it would explain the error persisting after the program exists or between restarts. The problem is clearly getting worse over time. It may be that there was a minor glitch initially and it persisted because of some sort of crap left behind which the scandisk somehow took care of, I suppose? But since then the problem recurred, only worse. Who knows.

Basically what I am wondering is:


1. Has anyone else experienced something like this? If so, any suggestions?
2. Does anyone know of a diagnostic I can run to figure out if it is my graphics card that is causing the problem? Unfortunately the people I live with either have laptops or don't have 3d cards, so I can't just swap in a known good card to see if that works. It's going to be a lot more of a pain to swap out the hard drive, so I'd like to start by testing the graphics card.

If any of you need any more information let me know and I'll do my best.

Thanks!

PS: I posted this on the Nvidia forums but got no responses... probably should have just posted here to begin with!

UPDATE: I Defragged my hard drive successfully -- except for a few files related to lotro and total war 2. My most fragmented files are data 1, 2, 3, and 4 for medieval, which are 1800, 1100, 1000, and 750 fragments respectively. Another medieval file is in 250 fragments. A few lotr related files are in 10-20 fragments, and a few unrelated items are below 4 fragments each.
February 25, 2007 3:05:23 PM

I'm running a Monarch rig that I bought a year and a half ago or so, back when Monarch was still reputable. Here are the specs:

Case: Thermaltake SViking
Power Supply: Antec True Power 550W
Mainboard: DFI Lanparty UT NForce4 Ultra-D
Processor: AMD Athlon 64 3700+
Memory: Corsair CMX1024-3200 PC320 (1x1gb)
Storage: WD 740GD 74gb SATA 10k Raptor
Video: EVGA nvida geforce 6800gt pci-e 256mb



I should be clear that I've only experienced these problems in relation to graphics intensive games. Other "heavy load" situations have been fine so far. Also, this isn't an issue of the system "dragging" -- it completely locks up when the bug occurs the vast majority of the time. In other cases it locks up temporarily, but the sound continues to stutter along at an extremely slow pace.

In the day or two between when the bug initially surfaced and now when the games were running, I noticed essentially the same level of performance that I've gotten since I got the machine, perhaps a small notch lower. It's pretty clear this is a binary situation-- when the glitch isn't active, the system runs just fine but when it is active, everything is fubared.

If you all need any other information please let me know!



UPDATE: I should also stress that these crashes are occuring on the *menu screens* for these games now. Not exactly heavy load situations. It can't handle the menu screen for med total war 2! It also can't handle the menu screen for Far Cry with the graphics settings previously set to "Low"
February 26, 2007 3:02:23 AM

What are your temp's like (both CPU and GPU)? It sounds like you could have some throttling going on.
February 26, 2007 4:06:22 AM

I'm with you on this one. I have a Geforce 6800 Ulta, and since BF2 came out, it worked fine and dandy. Now, play on maps with Jets and stuff, it will freeze then unfreeze and then i crash into the ground lool, but i can play INF maps fine. So I guess mines dying also ^_^
February 26, 2007 6:27:01 AM

Hmm, I would have suggested a format and reinstall of windows, but since you say its only happening in games then it must be the gfx card/ram. Try downgrading your display drivers to an 8x.xx/9x.xx (something thats not he latest). You can try that coolbits underclocking thing, but I doubt thats gonna do anything but lower some temps, cause the gfx card will still do the same thing. Thats about all I can say... I would still try a format if none of that worked.

Now I came to this topic cause I too have a 6800gt(agp), also giving problems. Your problems are completely different to mine though.
I would like some help from anyone that is able to....
I got my pc about a year ago :
6800gt
3.2 P4 Prescott
Msi 915 Neo-V PL
768 ddr400 kingmax(had 1gig corsair, but my pc was underperforming so I gave it to my cousin)
proline 300w with 15A on 12v+

The first problems I had was the pc not getting past the Bios, I could see it, it would detect everything and then freeze just before going to the next screen(verify dma data/ boot from), I fixed that by disabling Hyperthread.... why....I....don't...know.... but it worked so I didn't mind(gave away corsair ram).
Then came the other problem. The gfx card just wouldn't display. Before I go into more I have to say that my card is hot...like lava hot... it idles at 80C and constantly hits the threshhold of 120C in any game(2d/3d, old/new).
The problem is that the gfx card won't display for a few boots, I have to switch off the pc then switch it back on. Usually second boot it works, sometimes its stubborn and takes about 4, other times it just won't boot for days.
Seriously, the gfx just switches itself off for a day or 2, the pc sound does the usual 1long 2short for a gfx problem every time.
Now me and a friend tried to work this out, and ended up swapping gfx cards during one of those "no switching on days". The results were that my nvidia won't work in his pc, and his ati x800 won't work in mine. So my problem is the card and the mobo? I also have a broken pci geforce2 that displays ascii codes all over the screen, but it works every single time, I only use it to change bios settings whenever I reset the bios and the HT comes back on.

Recently the card only switches on once every day now... it still does everything fine when it works, plays all games(except RS:Vegas which runs a nice 12fps...), runs hot like normal....
If the card doesn't switch on I am still able to enter windows and do whatever I want(which is only listen to music).... just browsing blind with windows sam(I think thats his name).

I've tried just about everything.... including different mobo and psu(friends pc is running a 865 dfi lan party with 500w psu) and also flashing my bios. If anyone can help I would appreciate it.
February 26, 2007 10:16:31 AM

I have reformated my Computer few weeks ago [It's done this before and after that] i've used old and new video drivers. I have 2 gig ram, P4 3.0 I'm sure its just dying, only explains why I can play Infantry Only Maps perfect [cept like 64/64 and their all together in 1 spot lol] and jet maps where im in a jet/chopper.
February 26, 2007 12:54:07 PM

Thanks for the help all. The problems have been getting somewhat more severe. I tried testing the DirectDraw and Direct3D capabilities through dxdiag -- it managed the first few DirectDraw tests just fine but crashed when it tried the fullscreen bouncing box. Direct3D just crashed it right out. I also had one instance where it locked up on the desktop just after a restart, but that seems to be corrected for now.

As far as temps go:

Motherboard 38 °C (100 °F)
CPU 33 °C (91 °F)
North Bridge 52 °C (126 °F)
GPU 65 °C (149 °F)
GPU Ambient 51 °C (124 °F)
WDC WD740GD-00FLA0 22 °C (72 °F)


That's my current everest report, although I've got the side of the case off at the moment. My case temp monitor reads 22.7 degrees C. Typically with the side of the case on it was reading between 27-30.
February 26, 2007 2:01:59 PM

wow, it sounds like the 6800gt reached its predetermined date of death. Mine is also screwed up refusing to work with drivers installed.

All i can suggest atm is blowing all the dust out of it. It has previously been enough to fix my card, but not this time :( 
!