Sager/Clevo 8662 Nvidia gtx 260m GPU dying, PSOD, need advice

Recently been playing the new Star Wars game and I'm getting PSOD (pink screen of death) almost always instantly when I load up the game on the first level, and I believe from what I've read it's a VRAM issue? I never get this issue while browsing but I even get crashing on a low GPU game like League of Legends now too.
The laptop is almost 3yrs old and I only had 1yr warranty and only "install" would be warrantied. I now live in Thailand and it seems my only option is to look on Ebay. But now I'm reading that it may not be compatible with the Alienware or Asus 260m's which is apparently all that's on Ebay. Sager said a new Sager 260m replacement would cost me $470.. ex shipping, HA! I doubt I'd get that for this laptop with a working GPU.

Do I have any real options aside from searching for a new/used Sager 260m on Ebay for a reasonable price cause I'm SOL otherwise? I can't really afford to just go out and buy a new gaming laptop.

Added a few more photos

 

Have the same, but green instead of pink.
Bought the new laptop at the end.

M860ETU's time is passed away now.

 

im at work now but tonight im going to start a thread for everyone that has had a graphics card die. there are so many diffierent varients of card deaths that it would be a good idea to get everyone to post in one place to give owners a idea of what to look out for or what to try to rectify the problem..over the last year ive seen about 4 images similar to yours which as far as i can remember werent dead cards but cant find them as work computer is so slow.
will have a hunt tonight.

 

Not a hardware issue? oh please don't tease me. I'm fully emotionally committed to being doomed.
i added some more pics. i started to get the BSOD with the nv error then i'd crash out of swtor whenever i'd be tabbed out while the level was loading, then i started to get the PSOD then i formatted and i got it almost instantly upon entering a level of the game and then i was started to get a different 64 BSOD, where as before the format it just simply started to become more frequent. browsing however i got the PSOD once. damn i spent i think 2500CDN on this laptop 3yrs ago, i can't afford a new laptop. in a foreign country with no gaming? the past week has been the most boring of memory. worse than standing in the corner for getting in trouble when i was young.

oh and yes i took apart the GPU to clean and apply new thermal paste which the techs from sager said it sounded like a heating problem which i felt it wasn't judging from the crash logs from GPU-z. max my GPU had gone while this current testing was 70C from the logs.

 

You can always get a cheap card and throw it in

There are lots of "hidden" sellers around here with gems for these laptops. you can get a FX2700 for 80 bucks i think.

I am having some issues with my 280m gtx now. While playing guild wars, i get ALLOT of times a black screen and then it pops to windows saying that the "the nvidia driver bla bla bla but windows was able to recover". I hope my fate is now a dead 280m... nevertheless, i am finding a replacement just in case

EDIT: i do not have a M86 but a M570tu

 

download who crashed (in my sig below) which will give a much better dump report when you have a bsod.

 

The problem with color BSOD (pink, green) that it's not creating error/dump.
System thinks that all fine and that user himself rebooting host w/o a reason, based on win logs.

 

might well be worth trying a different driver, pref a whql

 

I have tried 290.53 (BETA), 285.62 (WHQL), 280.26 (WHQL) and 275.33 (WHQL).

Each time with total wiping out previous drivers. Result not positive. Effect is same all the time.

Temps are ok. It's definitely not overheating.

I'm totally wiped out system, low level format for both HDD, reinstalling win7 once again, from scratch, will try later with games which had 100% crash-color to check it.

 

On Sat 2/11/2012 9:59:28 AM GMT your computer crashed
crash dump file: C:\Windows\Minidump\021112-36239-01.dmp
This was probably caused by the following module: nvoclk64.sys (nvoclk64+0x1AB7)
Bugcheck code: 0x3B (0xC0000096, 0xFFFFF88008A46AB7, 0xFFFFF88004058E10, 0x0)
Error: SYSTEM_SERVICE_EXCEPTION

nice lil tool. thanks. this is the error that showed up. suppose the nice thing about this is apparently the PSOD i was getting was the same error that the BSOD was showing, so i'm able to feel confident on pinpointing the problem. out of 10-20 theads i've read so far, no solution. this has also like built up from youtube vids starting to crash, then crashing from loading screen from the game while tabbed out to BSOD'd then PSOD'd. then format and still BSODdddd/PSODddddddd! the only thing that's giving me hope is out of all these threads is i haven't seen someone say the solution was RMA lol.

h0w: you're still trying to figure out your problem? I don't get any problems until i load up any game, then once the graphics appear i got about 5-10s till i crash.

 

you could always try a RMA i couldnt resist.

this is a long shot but i found the most stable nvidia driver ever was 186.81
as youve spent so much time trying different drivers it might be worth giving this older one a go. if you still get pinky then its 100% not a driver problem.


186.81 WHQL Notebook Drivers

Release Date 27/08/09

For Windows Vista & Windows 7

* GeForce 8M, 9M, 100M, and 200M-series notebook GPUs.
* Quadro NVS-series notebook GPUs.
* Quadro FX-series notebook GPUs.


32bit NVIDIA DRIVERS 186.81 WHQL


64bit NVIDIA DRIVERS 186.81 WHQL

also could you post your full system spec.
2 questions. are you running sli and have you got a program called ntune loaded. just these 2 things keep popping up on google with the nvoclk64.sys error.

edit:
forgot to ask, when you do a clean install do you use driver sweeper,driver cleaner pro or cc cleaner. ive experimented in the past and whichever one ive used the others will still find more old nvidia crap on your system so i usually use 2 of the programs one after the other in safe mode and then all the old remnants are removed.

 

You could always try the oven I guess.

 

didn't see the 2nd page. I installed the 186 drivers still same error.
when i went to check i saw that this crash and the dump file dates didn't match up. so now i remember... i did install ntune cause i read that underclocking the shader clock could help. when installed ntune and i went into nvida gpu properties it would give me that BSOD crash. so i uninstalled and i never get the BSOD again but i get the patterned glitchy screen, i dont' even konw what the hell to call it?? i thought those dump files from the BSOD and the glitchy screen were the same, apparently not. thanks for helping me clear that up MrDJ you've been really helpful.

GTX 280 Freezes and Artifacts
this thread shows pretty much exactly what i get.

when i did a fresh install i do the same as i always do just a new partition format.

just a single card setup. gtx 260m



windows version: Windows 7 Service Pack 1, 6.1, build: 7601
windows dir: C:\Windows
CPU: GenuineIntel Intel(R) Core(TM)2 Duo CPU P9600 @ 2.66GHz Intel586, level: 6
2 logical processors, active mask: 3
RAM: 4289593344 total
VM: 2147352576, free: 1921478656

 

your welcome mate,thats what we are here for.

so one thing sorted. i didnt even know what ntune was but it kept popping up in google search with that error so thought it worth to mention.
as LaptopNut said a rebake could be an option but weve got to make sure its defo a dying card.
these are just guesses now but it might be worth getting some thermal paste and reseating the graphics card.
im going to try and hunt down some of the other posts with the same sort of paterns on screen like yours but this could take ages.

 

only 2 posts i could find were these buy they are different from your problem

http://forum.notebookreview.com/sag...lock-up-then-i-go-hard-reboot-every-game.html

http://forum.notebookreview.com/sager-clevo/641971-np-5753-video-card-failing.html

 

yeah i suppose those are close to mine... so if i decide to throw it in the oven you throw the whole thing in including the breadboard? i'd want this to be hella last resort. if i permafuk it... i'll have to consider going baack to canada to get a job or sell my body in downtown bangkok. i'm terrified of both.. i can't just stop playing games, it's not in my DNA

 

just in case here are the various methods of baking:

http://forum.notebookreview.com/ali...r-your-dead-graphics-card-your-alienware.html

http://forum.notebookreview.com/gaming-software-graphics-cards/385857-baking-gpus-oven.html

http://forum.notebookreview.com/har...-kind-guide-how-bake-your-gpu-fun-profit.html

http://forum.notebookreview.com/hp-...broken-geforce-go6150-fixed-after-baking.html

 

i'm reading good results from baking but what kinda risk involved? i don't have a backup nor could get one

 

so far everyone ive spoken to on here that has tried it it has worked for them. dont forget that this is only a temporary fix though so if you have no backup it might last a month or 6 months or even longer.
all it does is slightly melts the solder so the cracked links melt back together.

dont try what a student done in his dorm and use a microwave has he didnt have an oven. he didnt have a dorm after that lol

 

And to think they're handing these people degrees.

I'm going to toss in my vote for baking just for the heck of it. Don't bake your mainboard though. That's too much of a pain, and your mainboard isn't going to suffer from the heat issue that your GPU will, obviously.

If that doesn't work, call Sager. They might have a refurb for cheap.

 

so basically grab some tinfoil put it on the edges to raise it off the oven bars, set it to around 200C for 10mins. i don't need to remove anything off the card either. simple as that?

 

I would recommend going to the hardware store and buying a bubble level. Put it on the GPU, and use the aluminum foil to get it as level as possible. This might not make a difference, but it can't hurt either.

Also, remove any stickers or decals on the board as well.

Order some new Thermal Pads for you memory too. Can't hurt.

 

i live on the beach, i'm lucky to find chinese thermal paste. thermal pads? bubble level? lol. i have to drive 30mins on motorcycle to get tinfoil. i'm very lucky my place has an oven. will report back, thanks for your help guys.

 

Sounds/looks to me like either, a: your chip has burned up (thermal degradation). or, b: solder has became brittle below the chip.

 

You live on the beach playing on your laptop? Life must suck hard for you.

 

200C won't do the trick. You need to hit between 250 and 260C in order to melt the solder. A normal solder reflow that these things go through at the factory happens over a period of between 6 and 8 minutes. The temperature slowly ramps up over the first 4 to 5 minutes, hits over 250 for 1 minutes (the point where solder turns into a liquid and then cools down for the remaining period. You can emulate that by tweaking the temperature on your oven. A toaster oven is actually a little easier to control.

You don't want to heat it up too quickly or you'll thermal shock it and something will break.

One thing that I don't see mentioned in any of those links is the moisture sensitively level of those chips. Most of the silicon on these boards is MSL3, meaning that the package of the chip absorbs moisture over time. An MSL3 device absorbs enough moisture in 168 hours that it will crack the die when heated for a reflow after that. Basically the little bits of water that get into the package expand rapidly when heated and 'popcorn'... destroying bits inside the chip. The reels/trays are held in a temperature/moisture controlled area on the factory floor and everything is logged to ensure that they don't get screwed. Once chips have been exposed to ordinary air for long enough, they need to be baked to get the moisture out and reset the clock. That means cooking the chips for 24 hours at 125C.

Any PCB that needs to be reflowed after it has left the factory needs to go through that baking cycle or you run a very high risk of damaging the chips due to that moisture expansion.

So, what does that mean? If you are going to toss this board in an oven, you're better off cooking the board at 125C for a few hours at the very least to minimize the problem.

 

VERY good advice and a really nice explanation! I do have one question though, at the start of your posting you stated that they hit 250-260C to melt it but it slowly ramps up and only hits that for the last minute or so. Did I follow that correctly? So is your suggestion to cook that board at 125C for a few hours first and THEN ramp up the temp gradually for the 6-8 minutes to that 250-260C for a minute or two then drop it back down?

 

Yes, that would be the safest way to do it. The 125C cook to get moisture out of the package and then the 6 to 8 minute reflow to get everything back in shape. The 6 to 8 minutes includes the ramp up time and the cool down time... 2/3rds of the time is heat up, 1/3 is cool down roughly.

You don't need to do the bake and reflow immediately after each other, but the sooner the better of course.

A quick search on google will show what a typical reflow profile looks like for RoHS (lead-free) devices. The closer you get to it the better, but its not the end of the world to deviate a bit. Two real simple rules : Don't heat the parts up too quickly or cool them down too fast and don't exceed 260C.

A couple of hits from google on 'reflow profile':
http://www.token.com.tw/image/typical-rohs-reflow-profile.gif
Pb-Free Reflow Profile

The second one is interesting as it shows the MSL rating and how long the devices can be out of the package/dry storage. I haven't had to deal with anything MSL5 and above happily!

You can assume that anything in a BGA or QFN package these days is MSL3 as a rule of thumb.

 

60c seems to be quite a bit higher than the 200c recommended however that's good info also with the moisture. If i'm going to do this i'll want to do it correctly so ill ponder this.
The most important thing is for me not to fry or damage this board. If it breaks... well imagine being isolated from the world and your test goes wrong and you lose your laptop basically. I'm waiting for the Clevo P150EM (2-3 months away) and I'd hate to have to be forced into buying something i don't want. As much as I'd like to test this right away the risk vs reward isn't worth it when i thought about it. I feel bad cause you guys have gave me such great info and i want to deliver :\
it's actually forcing me to learn HTML and make web pages so i can make some money to afford a new laptop since my sugar momma hates me playing games and wont' replace it my laptop these cards are just so rare...! Unless I'm convinced it's quite safe

Doing a bit more research what about using something like eGPU ViDock? I could easily find a cheap desktop GPU in bangkok. Considering this laptop 260m is 3yrs old I'm sure it'd be much faster just for a cheap one.

 

I just found this in my /documents/conflict.txt

 

I have nothing like this in mine Docs.

Is this made by some 3-rd party tool?

 

did a quick search can't find anything.
doesn't this look bad? sharing memory resources and IRQ ports like that


all resources are shared in the box. the PCI-e is not where my GPU card is plugged in...

 

Hey WaryWW, I have the exact same problem as you which just started for me recently: PSOD whenever I play games, and sometimes even when the laptop is under no load.

I've gotten a single BSOD which looked exactly as the one if the first post. I updated drivers and everything, and still have the PSOD problem.

Basically, exact same problem. I've already tried everything suggested in this thread, and still no luck. I have not tried "baking" it yet though... which I'm a little scared to try.

Have you fixed it? or what action did you take? I called Sager technical support but they just said to ship it to them and I would be charged 80$ + part replacement for repair... which sucks.

I have a Sager NP8662, nvidia geforce gtx 260m, intel core 2 due T9900 @ 3.06GHz, 4GB RAM, Windows 7 Ultimate 64-bit

 

Same thing happened to me several years back. I will NEVER choose Nvidia over AMD again. And no, I really don't blame Sager for not covering the cost of Nvidia's inequities. F-Nvidia.

 

18 month old post

its just pure luck.

my m860tu is over 6 years old now and still going strong.
yes nvidia had a duff card with the 9800GT and as i always advise, if your going to spend a thousand or more on a lappy then whats another hundred on extended warranty. when my 9800 died it was changed the same day with a 260.
me never heard of many problems with the 260gtx.

even after all that i would never touch amd cards so good luck.