Sager 8150 - Screenflickering + GPU driver crash - help

This is the message i get after everything goes pixelated and the screen blacks out for a second.
ImageShack® - Online Photo and Video Hosting

Its happening intermittently even as i am typing this! I am using the 280.26 drivers but I've been using them since day one and had no issues.

Everything was fine until about just half an hour ago

My temps over the last thirty minutes:
ImageShack® - Online Photo and Video Hosting

What could be the problem? I havent started trouble shooting yet. But Im going to try and do a clean reinstall of the Nvidia drivers first

 

Just reinstalled 280.26 drivers. It was stable for 5 minutes and then its back to crashing and pixelated flickering again.

I have never done OC on this and temps were always kept below 80, using Fn+1 fan speed.

 

Having a similar problem with my laptop, only happens every hour or so, but still rather frustration, I'm using 280.26.

 

Your temperature look like mine? nothing is above 70c

 

that is a very old driver and might not officially support the 485 gpu. or is it a typo and you meant 280.26

you can get the last 10 drivers from > NVIDIA Driver Downloads - Advanced Search just to test if its actually the driver causing the problem.
if you update to a different driver you might need to do a manual clean in safe mode to get rid of all the old remnance.
if you need a walkthrough just holler.

temperatures are fine by the way.

 

It was a typo. I am using the latest driver, 280.26.

I just captured a video showing the problem, to send to Sager and to post here for the benefit of others having the issue. Will post back once its uploaded to youtube.

I had this same thing happen on my 8600M GS on the HP Dv9000, three times, twice inside the warranty and then outside of it.

I am really really surprised this happened to my 485M on the SAGER!. Its just been about 3 months of gaming with it and I always run games with fan running at max speeds to keep it cool. I have never tried OCing either.

Just my luck!

MrDJ, is it not enough to just check the "clean install" option when reinstalling the Nvidia drivers? I do not know why we have to go into safe mode to do that.

 

Hi, let us know what sager model yours is and when you got it. Every hour or so is still bad. I can reproduce the issue if i open any video, flash on the browser and even something that takes up graphics on screen.

Also, could this problem be due to a bad thermal pasting job by Sager? But the temps have never spike. Is there anyway to detect the effects of bad pasting?

 

If temps aren't spiking, it's probably not a bad paste job. The only way it could be without temps jumping would be if there was so much excess paste that it shorted traces on the motherboard. A graphics glitch would be a surprisingly minor symptom if that was the case though

 

Unrealted to thread: Could bad thermal jobs produce CPU/GPU throlling?

 

Sager 8150 NVidia 485M blackouts - YouTube video about the issue.

Funnily, I am unable to reproduce the crash anymore. I was running the laptop for three days straight and it was the third night (yesterday) that the problem appeared.

After that, i shutdown the laptop and let it rest for the night and today I am unable to reproduce the issue with ANYTHING!!!! I didn't do any changes. Just backup'd my files and stuff.

Mystified!

 

there are arguments for and against this. until nvidia updated ther driver package for a auto detect install the safe mode option has been used as far back as i can remember and has never caused me problems.
in safe mode you run driver cleaner pro or driver sweeper and lately i have been using cc cleaner as well and it always finds obsolete files from nvidia that the normal instal just does not remove.
OEM's recommended this option for install and thats why i still use it and recommend it.

 

Thanks for the background info. I'll give driversweeper a shot. Wouldn't hurt to try.

But now, I haven't had the crash all day and I've been running Nvidia graphics demos, running videos, flash etc. to try to reproduce the crash.
Yesterday, I had the crashes even after a "clean install" using the Nvidia installer.

 

my saying is if it aint broke dont try and fix it.
if its been behaving today see how long it lasts before it goes again.
it could even of been a rogue windows update that has now been fixed.

ive never used driver sweeper but others have.

just in case you want a step by step guide for the future:

 

Thanks for the advice and information. Unfortunately, I cant seem to give you rep. I'm guessing you were the last person I had +1 some weeks back, so the popup says "spread some rep around before you rep this user again!"

Edit: I am not trying to fix it anymore. Ill wait for the next crash if it comes.

 

no probs mate. as long as it helps.

 

Here's a repost of a topic i made on gamefaqs pc board...

Used to run Crysis maxed out at 720p with 30 fps solid, now its slower!
wheepitup
Posted 9/4/2011 3:56:51 AM
message detail
quote

Sager np8170 - Overlook and Crysis Warhead Benchmark - YouTube

A vid showing how Crysis Warhead used to run.

Now i play the same part and i get a mix of 24-40 fps, it seems really unstable. In Crysis 1 the framerate was dropping as like as 12-18 fps. Something weird is going on with my GPU. I have the newest Verde drivers and all. I wonder if the newer drivers are somehow causing this or my hardware is just crapping out.

GTX 485m
i7 2630qm
4 gb ram

I did disk defrag, error checking, tried rolling back the nvidia drivers. Nothing fixes it. On MSI afterburner i'm getting high temps of 98 degrees.

So yea, lately i've been getting strange performance dips and increases, and my temps seem kinda high (is 98 degrees high?) I tried rolling back my gpu drivers to the last official verde release and it didnt help, so who knows.

 

Uh yeah! 98C is INSANE!

Mr. Mysterious

 

That's under load with Crysis 1 maxed out at 720p

 

Lol decayedmatter...it's your computer. Do with it what you wish...but those kinds of temperatures honestly scare me.

Mr. Mysterious

 

Well i just tried crysis 1 again, and it ran at 94 degrees but the framerate just doesn't feel the same as when i first bought it!

 

Hey decayedmatter. That temp is way too high, isn't it? I was increasing in my temps for CPU and GPU gradually over the last few months. I mean that, when i first got the laptop the idle GPU temp used to be 39C and the 4 cores at 45-. Now, they are at 43 and 50 respectively. So yesterday, I opened up the laptop and boy, I found a clump of black dust on the heat sink and got rid of it. Cleaned the fans too. Now the temps are back to normal.

EDIT: I always use Fn+1 for crysis 1 and 2 and keep the temps at 86 below.

I mean cleaning up the fan could fix your FPS drop issue! But relatedly, in this Nvidia forum http://forums.nvidia.com/index.php?showtopic=207666&st=240 a few of those who have experienced driver crashes like me say that even after rolling back to 275.33 (dont know if all of them did the clean install) they keep experiencing crashes. One or two of them even said that a clean W7 install + 275.33 didnt get rid of the issue. So, permanent hardware damage from the new driver set?

 

Yea i just sprayed some compressed ait into it but the can ran out, i think ill open it up and see if theres any dust clumps. You just remove the far back plate right?

 

Yeah, on the 8150, it was just four screws to get the back plate (the one with the subwoofer grill) off. and then 4 tiny screws for each fan cover-plate.

The fan blades themselves just slip off, so i took them out and cleaned them good. I didnt remove the heat sink assembly itself, but just used a brush and compressed air to clean the fins.

 

I opened it up and took a look, but the fans are incased inside the heat sink, i would have to remove more screws to get the fans out. So i didn't do it yet since i need some compressed air first.

But from what i could see, there's hardly any dust on them, just a veeeeerrry thin layer on the blades, there was a few hairs stuck to the heat sink so i removed them with tweezers, but right now i got my laptop going with max fan speed and its at 68 degrees, seems kinda high for idle temps. IT's gotta be either an issue with the nvidia drivers or my gpu is crappin out, cause i didn't see any major dust buildup to cause this.

In fact check out this quote from me when i first got the laptop...

Something is definitely off. Think i should contact Xotic PC? Or will they not care till the laptop stops working?

 

Something is definitely weird with you temps. The 75 that you got when the machine was new is amazing. The 98 now is just so dangerous.

Coming back to the fan issue first, I didn't remove the fan-assembly entirely nor did i have to remove the heat sink assembly.

Each fan-assembly has a metal plate on top of the blades right? That metal plate is held in place by four tiny screws. After removing the plates, you can access the fan-blades and remove them just by pulling. And THEN you will see the copper insides of the heatsink-fins where the dust actually is! EDITNOTE: the fans were not encased in the heatsink assembly. Only when you remove the metal plates, will you realize this.

I had a lot of dust coating the fan blades though and dust clumps on the inside of the heatsink-fins. Wouldn't hurt to clean it thoroughly and trying out the temps again before calling xotic. You would only need 10 minutes for all of this.

If there really is no dust and your temps are so bad, perhaps its the drivers? You said you already rolled back to 275.33? Did you do the clean install? MrDJ has posted the walkthrough for a really clean driver install on this thread. You could try that too and post back here.

I don't know much about the thermal pasting issues but if it was done badly then it should have shown up on week one itself, right? So, I dont think its that.

So far all the ranting about the problems with 280.26 drivers 280.26 Display driver nvlddmkm stopped responding and has successfully recovered. - NVIDIA Forums on this Nvidia thread have said nothing about temperature issues.

 

I think i figured it out but something is still weird. I have been running dual monitors so i turned that off. I ran Crysis again maxed out at 720p, got 33-40 fps and temps only went up to 61 degrees!

So i'm thinking, ok...it was obviously my dual monitor setup that was causing these high temps, since now my idle temp was 44 degrees.

So i boot up Deus Ex Human Revolution, run it at 720p again, with dx11 on and everything maxed out. HOWEVER, now i'm getting a load temp of up to 94 degrees!

Sure thats not 98 degrees like before, and when i exit the game it goes back down to low 40s idle temp.

I dont know if its the tessellation and advanced depth of field in Deus Ex or what, but it seems weird cause i get 60 fps solid in Deus EX but tons of heat, and only 30-40 fps in Crysis but only 61 degrees of heat.

I should also note that i rolled back my verde drivers to 275.33.

This is SO CONFUSING! I'm gonna try a few more games right now like Crysis 2 in DX11 and Metro 2033. Give me a bit.

 

Haha I knew that you use an external monitor because you said so once somewhere on NBR.

I have the Alienware Optx AW2310 3d monitor and running it on dual monitor setup makes the idle temps go to 58C. Running just the external monitor = running just the laptop panel = 40C idle temp.

But when i game, I make sure that the laptop monitor is off. I assumed that you would be doing the same and that in-spite of this the temps went to 98.

Whats up with deus ex. Did you try playing with the power management option in the NV CPanel? I keep it on "performance"

 

Ok, i don't know is going on but i doubt it has anything to do with your topic anymore. I just tried Crysis again, same spot as before, this time my framerate was 20-25 fps and temps in the high 90s.

When i turned off dual screen, then played crysis, i was getting up into the 40s for framerate and temps stayed at 60 degrees. Now its all screwed up again!

Ughhh, this is really worrying me now.

 

Well dual monitors does put a bigger load on the GPU and will also cause FPS to drop.

Dual monitors and playing a game still should not cause a GPU to hit mid 90's but is normal to see poor FPS with them.

In windows for example one monitor your GPU will downclock to lower Mhz when idle and thus lower temps. If you have dual monitors it does not downclock and will run at full speed, even if idle. So that causes idle temps to be higher.

 

I figured it out guys, i opened up the bottom again but this time removed the cpu and gpu fans, turns out the entire inside of the vents leading to the back of the laptop were CLOGGED with dust. No clue how it got that dusty since it's always on a smooth and clean table. But nonetheless, i cleaned it out good and now i get 34 degrees idle and below 70 degrees under load.

Thanks for telling me to do that mountainlifter! Now my only concern is if i damaged my gpu and cpu running at 98 degrees for what mustve been at least a month.

 

Well, if you are like me, you will feel that deep inside you know its life "may" have been shortened, but at the same time acknowledge that it "shouldn't" because the cut off is 100c which is what Nvidia uses for cut off before perma damaged is caused. But I think anything running hot effects its life, even if within spec.

However, don't that scare you. It very well may last 10 years and still be kicking

Also on the dust part, if you run Fn+1 that will cause dust to get in her more often because it has such massive amounts of air flow when on high, you don't think about minor dust or even hair getting in it, regardless of how clean your flat desk may be. I on on a glass top desk and I can see the dirt and dust very easy hehe

We are all glad to hear that you figured it out though always good to find a solution that does not require an RMA. After my RMA hell, I feel for anyone to ever want to do one...

 

Ughhh yea, i'm still a bit worried but i didn't ever exceed 100 degrees. I hope no damage has been done. Definitely will keep my eye on the temps more often from now on. As for mountainlifter. I hope he solves his problem as well. I have no clue whats causing it for him though.

 

So it was the dust! Great news and glad i could help. My problem is actually mental.

Day before yesterday, when i sat down to work at 8pm, it (GPU Driver) just kept crashing over and over and over. I called my canon-dslr pal over and made him take this footage of the repeated crashing. Sager 8150 NVidia 485M (280.26 driver) blackouts - YouTube

I clean installed 280.26 again but the problem remained. I uninstalled 2 software (stardock fences and dexpot), backed up all my files, and got ready to RMA cause i was sure the GPU was bust. The same symptoms I had seen with my 8600MGS on my old HP laptop twice, and the GPU (ie. the entire mobo) was replaced both times.

I shutdown the sager that satuday night. BUT, yesterday and today there have NO CRASHES. Now, i'm trying to reproduce the crash in many ways. HAHA. If it wasnt for that youtube video, i'd have thought i dreamt the whole thing up.

 

From that video, it seemed like Windows was switching between Areo and Basic for no reason. Windows is supposed to do that only when the system gets overloaded, but windows has a really paranoid definition of "overloaded." For example, if I run a game and forget to snooze BOINC, my computer will do that sometimes. If that happens again I'd try temporarily switching to Basic in the control panel and see if the switching still happens.

 

It is not just the switching from aero to basic. In fact, that was hardly the issue. Opening videos, flash on chrome etc. crashes the driver with the message "Display driver stopped responding and has recovered.
Display driver NVIDIA kernel mode driver, Version 280.26 stopped responding and has successfully recovered"
I dont think this TDR message pops up when windows auto switches from aero to basic, or does it?

 

No, it doesn't- but the reverse is true. If that message pops up, it usually goes from aero to basic until it recovers. Have you tried uninstalling and reinstalling flash? There was a release not long ago that had similar effects on Nvidia drivers- though the fact that this is happening outside flash applications or browsing suggests that it's something else entirely.

 

Oh, so that's what that message said. I couldn't read it from the video.

 

I'm getting the same thing on a 8130 with 460 GTX. I don't think it's the temperature since every time the driver crashes it doesn't feel like it's blowing hot air. This happens to me every 3 or 4 days. All I can do to fix it is shut down the computer by holding the power button and restart.

-B

 

Thought I had my specs in my sig. I reinstalled 280.26 and it had no effect sadly. Just went black not even a few minutes ago. Hope your laptop keeps working.

Just another though, anyone have any idea why I can't seem to use the 275.33 drivers? Whenever I run the installation program the program just says "The graphics driver could not find compatible hardware on your system."

MALIBAL Lotus P151HM1
| 15.6" FHD LED Backlit Matte | Intel Core i7 2630QM | NVIDIA GTX 560M | 8GB 1333MHz | 500GB HDD

 

The 275.33 drivers may have been before Nvidia added Clevo support to the driver files. Check out the thread below for a modded inf and you should be able to get them running from there

http://forum.notebookreview.com/sager-clevo/564375-atmosks-nvidia-clevo-support-modded-infs.html

 

I had put up the TDR message in the video description.

Yeah, the temps were fine in my laptop even when it kept crashing over and over. You didn't mention your driver version. The feeling is that its the 280.26 drivers. I suggest following this thread: 280.26 Display driver nvlddmkm stopped responding and has successfully recovered. - NVIDIA Forums Perhaps you could revert to 275.33.

No crashes so far for me.