Can someone tell me what the hell is going on?

First; here are my specs:
Sager NP5796 / M570ETU
QX9300 @ 2.93Ghz
8800M GTX
4GB RAM

Okay, this problem actually started long ago, but has gotten worse in the last two weeks.

I noticed when playing GTA IV, that sometime the game would just freeze for maybe 5 or 6 seconds. Then it would run perfectly normal again. When I looked in the RivaTuner-log for videomemory, I saw that videomemory had been dumped when the game froze.

These weird freezes would happen maybe 2-3 hours apart, but eventually they started happening almost every 30 minutes. They usually started occuring when I had played for maybe 3-4 hours. I thought maybe it was a memory leak or something and didn't pay it much attention. No big deal I thought.

Fast forward a month, and I started getting more and more error messages about "nv4_disp has stopped working normally and you must reboot your system". These messages popped up in a normal windows pop up box. Also, along with this message the dekstop resolution would drop down to 640x480 and over to 16 bit. These errors I got maybe once every week or so. No big deal I thought.

Fast forward to today. For the recent weeks I've only been experiencing the nv4_disp error maybe once a week. I haven't had time to play any games.

So today, for the first time in 2-3 weeks I had time to play some games. So I started GTA IV. Immidiatly after I went out of the safehouse, I noticed a strange FPS drop that has never happened before. The FPS got a bit better after 20 seconds or so. No big deal I thought, and stole a car. After driving around for 30 seconds, the game freezes. And some weird happened, a texture like this filled the entire screen. The radar was completely borked, it turned into a black rectangle with some blue stuff here and there. Game froze completely.

So I used ctrl+alt+del to terminate the game. After 10 seconds on the desktop, my mouse stopped on the screen and everything became unresponsive. So I had to do a hard reboot by holding the powerbutton for 5 seconds.

I tried GTA IV again. Same thing happened, except this time a texture of a brick wall filled the screen. This time, pressing ctrl+alt+del didn't help so I rebooted.

I've tried the game 10 times now. Same result every time. First some weird FPS drop after I go outside the safehouse. Then the game either shows a random texture on the whole screen, or just freezes suddenly. Sometimes ctrl+alt+del works, sometimes not.

I wanted to play a game, so I fired up Battlefield 2. Everything seemed fine for the first 10 seconds. Then all textures suddenly turned to a mixture of black, purple, yellow, you name it. The text on the screen also became borked completely. Crosshair turned into 4 red, ugly rectangles. See [ http://img21.imageshack.us/img21/9520/32qw3.jpg
However, the hand and weapon model in front of you, and models of other people around the map, seemed fine. Game also ran fine, no FPS drops or whatever. But I can't play with these ed textures.

So I tried COD4 Singleplayer. PC locks up during the introduction where the guy talks about Soap.

3 games, all same results. Freezing, textures ing up, sometimes ctrl+alt+del works, sometimes don't. And sometimes even if I manage to terminate the game, the desktop will freeze 10 seconds after I've terminated it.

Let me tell you that I tried 2 different drivers (newest and 179.48), and all had the same result. Re-installing drivers does not fix this issue. I always used safemode+driver cleaner and driver sweeper to remove drivers completely.

The temps of my CPU are 89-98 under load. The temps of GPU can reach 95 degrees under load.

I know those are high temps. I tried cleaning dust. Temps decreased a bit, but problem still remains. What's going on here???

 

they are both at the threashold and they're prob stepping down to prevent meltdown

thats wierd... maybe you accidentally activated silent mode

hit FN+2 for the fans to go on full blast and try again

 

The normal temperature should be <90C so its obviously an overheating problem, may be due to a heat sink malfunction. Where are you using the laptop is getting some air, right? Have try using a cooling pad? Have you opened up the laptop and see if anything is blocking the vents? The body has the usual vents? What os are you using? win7 gets hotter than xp.

 

In BF2, GPU only reached 70 degrees before I got this

CPU was about 40 degrees at the time.

I don't really think it's an overheating problem...? I mean, if it was, weren't the textures supposed to screw up when the GPU/CPU reached a hotter temp. This time they screwed up at very low temp.

EDIT: Right now I'm running RTHDRBL. Temps of the GPU are currently 92 degrees, but no errors thus far. The ATiTool test utility also shows no errors.

It's just in those games.....

 

did you try fn+2 to run the fans in full blast

 

Yup, I always keep my fans at full, and I sit next to the laptop so I can definitly tell the fans were running. It sounds like a freaking jet engine after all...

EDIT: I just exited RTHDRBL, and the computer hung for 30 seconds...mouse and everything froze, but then it came back to life again. What the hell is this crap?

 

Here's another fun screenshot of BF2: http://img408.imageshack.us/img408/4378/mhmfh.jpg

 

Your temps are clearly unhealthy.

Apart from that, I would format, re-install, update drivers.

 

That looks like memories going bad. Have you been overclocking them ? And yes, those temps are too high. Should be below 80C.

 

Any overclocking in the past??

 

Yah if it kept getting worse and worse over that peroid of time, it sounds like a hardware issue. You can reinstall your OS, all drivers then see what happens, if nothing changes, you'll have to consider a hardware failure. nv4 is the video, so it sounds like the video card is acting up. You could always try baking it in the oven to seal the soldering ( there is a guide on here ) if you are absolutely sure it's the vid card, and you have no warranty left.

 

Format; reinstall OS.

Clean CPU/GPU; reapply thermal paste.

 

I think its a sign of a video card going bad. My old 7900GS used to do the same thing (artifacting after 5 mins of Bioshock or similar games) till I had to get it changed.

Duane

 

I haven't been OC'ing the video card ever. Only the CPU.

I ran some tests, and GTA IV seems to crash when the GPU reaches 78 degrees, which is not particularly high. The CPU was at about 50 degrees. (All in celcius)

Thanks, I'll try that too.

Did you get the same effects that you see in my screenshots?
When I hear the word artifact, I think more of this, rather than this, which is what I am experiencing. Notice that only the textures are ed. No random lines or anything like that. Also the FPS is steady and high. As one of you said, maybe the videomemory on the card is faulty. Any way to test videocard memory?

Well, I've tried reinstalling drivers. Not the OS though, cause it's a pain in the and a process that takes time. I want to rule out every other possibility before I try something drastic like formating everything.

Anyways, I find it dissapointing that this 8800M GTX-card has only lasted me 1,5 years. One would expect much better lifespan of such an expensive electronic product if you ask me. I've been taking very good care of my laptop, and this is what I get. First thing that failed in my last laptop, was also the graphics card, of course a nVidia card. Seems nVidia products are just crap. I swear to God, I will never buy a nVidia product again, unless it comes with the laptop I purchase. I've had nothing but bad experience with nVidia, and it doesn't seem to stop.

Any chance of fitting an ATi Card in the laptop? I heard rumours of a 4870, but that might be for the newer generation, the WU87CU or whatever it's called.

Last thing I need now is having to replace the video card. As of the moment my economy is not at it's best, but I really need a computer for work.

 

http://forum.notebookreview.com/showpost.php?p=5397451&postcount=818

 

Sorry Cookie, looks like your memories are dying out. I suggest you stop gaming or any activity that heat up your memories otherwise it will get worse and you might not be able to use your laptop at all.

It's not your fault, stuff like this happen, it has happened to me twice, once on a 7900 GS (in warranty) and then on 7900 GTX (no warranty) with no OC -> which forced me to change my laptop to the one I have now.

The only thing you can do is as Soviet already wrote... test your memories and if they are bad... I guess your card will have to face the oven. I actually think that in your case the oven will solve your problem.

Remember Cookie, we support you in your endeavours.

 

Part of the reason as for why GPU memory dies quickly is because they aren't given the level of care that the GPU die gets all the time. What I mean is, the thermal pads that bridge heat from the chips to the heatsink isn't replaced at all in almost every notebook I've seen. I can bet you nobody in this thread has replaced their thermal pads on their GPU memory. Thermal pads will dry out and degrade over time, especially when the heatsink is often removed and remounted. Usually they should be replaced annually for high performance GPU's if periodic maintenance is done on the notebook. Mid tier and budget notebooks do not need to worry about replacement on an annual basis because the GPU heatsink is seldom removed. They can pretty much operate on the same thermal pads for the full lifespan of it's service, usually around three to four years or even longer.

For example, I was opening my client's notebook about a month ago, an older Dell with a 7600M GS. It has been opened before by the owner and by unauthorized repair centers and I can tell you it's been through hell and back; missing screws, electrical tape substitutes, too much thermal paste, poorly soldered leads, etc. But for the sake of relevance, the thermal pads were dusty, stiff, and were starting to tear, which are sure signs that it is well worn out already. The less cooling performance the thermal pads can give the memory means that the memory will be running hotter, and I think we all know what happens when memory gets hot: shorter lifespan and instability at lower GPU core temps.

I've read your previous posts, Cookie, and I can tell that you take very good care of your notebook just like the rest of us Clevo users, except for Heathkidd. I am not saying that you frequently remove your heatsink often, or poke and prod the thermal pads, or not clean them, rather I am saying that the possibility of damaged thermal pads is there. Only you would know if your thermal pads are wearing out.

Even if that was the case, your GPU is already damaged and the only options you have right now is to search for a replacement GPU for a reasonable price or bake your card.

 

Is your laptop still under warranty? Because if it isn't, it's worth giving the oven method a try...

A little off topic, but what's the best method for applying thermal paste to the GPU? Pea/Dot method?

 

Could somebody do me a favor?

Download this test: http://freestone-group.com/video-card-stability-test.htm

And tell me when you run the test (press start). The planet spins smooth on your computers? Cause on mine, it lags very much, and sometimes stops for 2-3 seconds. Seems there is quite a bit of lag.

Also in RTHDRBL, I get constant stutters that occurs exactly every 3 second. I can see from the FPS counter that FPS dips from 70-50 every 30 second.

I'm also starting to suspect it might be my RAM. Maybe I should do a memtest.

 

I downloaded the program and ran the test, it didnt lag at all on either of my systems (in sig).

 

Thanks

That means something is wrong with my GPU or my driver. What's weird is that when I try the Stability Test, I get no artifacts even if my GPU hits 96 degrees.

In games I don't get artifacts either but they all crash.
( Only game I get artifacts in is BF2)

 

Would inserting a different harddrive from another laptop into my laptop work? I'm thinking since my other laptop is an old Inspiron 9400. If there is still problems I'd be safe to say it's a hardware problem, because driver issues should be elimited with a completely different instalation, right?

 

In theory... yes!

Just use stock drivers to make sure all is alright.

 

Running BF2 in batterymode seemed to work fine.

I played for 10 minutes. No problems.

Checked the temps...58 degrees maximum.

But the moment I put the powerplug in: BAM, corrupted textures.

 

BUMP

10char

 

Cookie, can you take a picture of the artifacts and post it here?

 

Hi, thx for the reply.

I have two pics in the OP....the screenshot function seemed to also capture the artifacts (don't know if it's always this way with artifacts or not)

 

It seems that lowering the memory clock for the GPU in the nVidia control panel fixed the problem.

I lowered the memory clock from 799mhz to 720mhz and the crashes stopped occuring.

This is weird. My GPU can't handle the default memory clock speed? Has nVidia overclocked the card from factory or what?

 

Sounds like you might have faulty memory on the card. Do you OC with FSB at all?

 

Ok, I think he is right, your GDDR no longer wants to be nice. It will progressively get worse until you won't be able to play anything. I have experienced the same symptoms on my previous 7900 GTX Go. Baking the card might be a solution... I will give it 3-6 months more depending on how much gaming you are going to do.

Sorry

 

Oh well, as long as I only have to lower the voltage 60mhz...can't be too bad for performance.

When it get's worse I'll probably bake it.