evga 7800 gtx ko dead after a game crash?

[skinnypotbelly]

A couple of days ago I started having some problems with F.E.A.R. freezing my computer AFTER I quit the game. I looked on the F.E.A.R. forums and it seems that a couple of people had the same problem I did. The work around was to either: install a No CD crack or to leave v-sync enabled. I left v-sync enabled and everything seemed to be ok but I was only able to test it out a few more times since I was close to the end of the game. After this happened F.E.A.R. started to crash on me at startup too (twice to be exact).

Fast forward to a few hours ago I decided to play COD2, since I had beaten F.E.A.R. Within a minute of starting the game, it crashed and restarted my computer. After it powered up again I decided to defrag the harddrive and run check disk for errors just in case. On step 5 of check disk the computer froze again and I had to manually restart the computer. Since it froze, I decided to uninstall COD2 but halfway through the uninstall it crashed again. After it automatically restarted, it got the the evga splash screen but seemed to be "stuck".

I've never seen a computer crash during an uninstall of a program so I thought it might be that harddrive. So I decided to unplug that drive. Restarted but now I wasn't getting a signal to my monitor at all. Thought it might be my SLI setup so I uninstalled one card. This time it got to the evga splash screen but it seemed like it started to artifact? Little black and white dots started to appear on the screen. It froze on the evga screen like the other time too.

I switched out to the other card but that one wasn't sending an image to my computer either. Since then I've cleared the cmos and tried starting up with just my main hard drive, but nothing seems to work. I'm not getting any type of signal to my monitor. I'm pretty sure its not the monitor either because Dell has a simple diagnostic test to see if their monitors are broken and mine passes that test.

Looking at the post codes it usually gets stuck on C1. Which is memory presence.

Is this a video card problem? motherboard problem? hard drive problem? I might try to go to compusa tomorrow to get a video card to test out. Other than that I don't know what to do. Any help would be greatly appreciated.

My specs
amd x2 4800
evga nf4 sli mobo
2 x 7800 gtx ko
crucial 2 x 1gb ballistix
seasonic s12 600w
2 x wd raptor 74 gb (one for operating system, one for games)
1 x maxtor diamond max 300 gb
creative x-fi extreme music
cheapy compusa pci sata controller (says its silicon image sil3112)
dvd burner and a dvd player
7 fans (2 running off of the motherboard, the rest on a controller or directly connected to a molex pin)

Nothing was overclocked. I've tried to overclock the cards twice to run some benchies, but it was artifacting so I left them at stock. I've never seen the cards go past 78 C (on a hot day), so I don't think it was a heat problem.

 

[xtknight]

Yikes...I don't know man.

I say first unplug one of the video cards and limit it down to which one is artifacting/causing issues.

I do know that the Diamond Max 10 drives have issues with nForce 4 boards. I have had my DM10 250 GB not detect on me for as long as I could remember with my A8N-SLI Deluxe board. You'll want to avoid installing nForce4 ATA drivers or enabling NCQ to limit out hard disk problems, probably (in the future).

 

[skinnypotbelly]

Turns out it was my cheapy pci sata controller. I think all of the crashes caused the driver to be corrupted somehow. I've uninstalled and reinstalled the card and now everything seems to be fine. Just to be on the safe side I've updated all of my chipset drivers and video drivers. xtknight, do the ATA drivers in the new 6.70 nforce 4 chipset drivers still have that problem? I've been looking around on the net and most people have said that the 6.70's are pretty good.

 

[xtknight]

Yeah they do still have the problem. The SATA boot ROM on the DM10 drives is faulty but seems to work OK with most non-nForce4 mobos. You can still install the nForce4 drivers, just not the ATA part of it. There is really no reason to because it will cause you nothing but problems. I had to go through bunches of BIOS updates and it still wasn't fixed. Some nForce4 ATA drivers don't like your certain BIOS version...it's a mess. All this gets you is NCQ which is useless unless in very certain scenarios.

 

[skinnypotbelly]

Another update.

After I thought I fixed the problem last night I ran memtest86 just in case. Sure enough, came back an hour later and I had over 700 errors. I've ran memtest before on the same sticks (crucial 1gb ballistix), in the same slots (1 and 3) and it ran for 5+ hours, 6 passes. Reboot, and same thing happened again; it wasn't POSTing. Took out one stick, everything booted up normal. Switched sticks, but used the same ram slot, everything booted up normal. Tried it again in dual channel with the same slots (1 and 3) as before, wouldn't POST. So now I'm thinking that maybe ram slot 3 was bad. Switched slots to 2 and 4, everything was fine. Ran memtest86 overnight for 5+ hours, 6 passes. So I thought everything was ok.

I decided to play COD2 this morning and withing 5 minutes it crashed again. Rebooted by itself and windows reported that the system32 \ drivers \ ntfs.sys was corrupt or missing and it couldn't start windows. I took out one stick again and everything booted up normal. So now I'm thinking that one of my ram sticks are bad. I read online that some people were having that ntfs.sys corrupt or missing problem with a bad stick of ram so I decided to test each one separately.

Now this is the weird part. So far I've only had this crashing problem and it not being able to reboot with COD2. So I decided to test each stick separately with COD2. Using the same ram slot (2) each stick is able to play COD2 wiith no crashes. I played both (one stick at a time) for about 30 minutes each time. Before if I had both sticks installed it would crash within 5 minutes.

So is it my ram that is faulty? If it was would it be able to play COD2? I've ran memtest on both sticks at once and that passed with no errors. About a week ago I ran OCCT torture test for about 30 minutes because I loosened the timings a little bit and that passed with no problems.

Could two of my ram slots be broken? I only seem to have this problem when I'm running dual channel. The reason I ask is because I've ran it in slots 1 and 3, and it had problems as soon as I installed a stick in slot 3. So I swtiched the ram to slots 2 and 4. Although it passed memtest, it crashed on COD2 in dual channel, but testing each stick separately in single channel, everything was ok.

If COD2 was faulty would it even run at all? I might try to reinstall the game to see if that works but I don't think thats the problem because it did run with one stick of ram installed.

As I am typing this, I'm running the OCCT torture test on one stick of ram and its passing with no errors. I'm going to switch sticks and, if that passes, I'll test each stick separately again on different ram slots. Should I use another program to test the sticks? superPi? memtest test 5? If I use memtest test 5, how long should I let it go? Help, please.

 

[deathwalker]

I would be thinking about detuning the ram settings and bumping the ram voltage up in steps to see if the problem goes away.

 

[skinnypotbelly]

I think I narrowed it down to ram slots 3 and 4 malfunctioning. I tested each ram stick separately and each ram slot separately. I found that if I had a stick installed in slots 3 or 4 that the computer would either not POST or crash immediately after getting into windows. Slots 1 and 2 booted up fine and I checked out fine with OCCT and memtest.

 

[rapty]

I am currently haveing the exact same problem as you I have run all my ram through memtest and all seems ok in ram slots 1 & 2 I get the problems with freezeing in games when I run 4 sticks.

how can you tell for definate that its 3 & 4 malfunctioning?

also what is your fps and score in 3d max is it slower than normal?