asus n56vz crash

hi,

I've got new asus n56vz, installed win7 ultimate x64 and there appears a problem, when playing games. Here is a crash photo: http://img21.imageshack.us/img21/3996/3ia3.jpg
I don't know why, but it depends on game, for example, I can play Diablo 3 for hours and maybe this screen occurred once while seeking max temperatures was: CPU 85, GPU 70 degrees Celsius. Meanwhile playing Civilization V it occurs almost every time after some time playing and max temperatures are CPU 75, GPU 72 degrees Celsius. The only option then is to force shut down. Any ideas and solution how to avoid this?? Thank you very much.

 

Ouch. That looks like a dying video card. Usually that sort of issue is caused by bad vRAM.

If you like, you can try testing the video memory (vRAM) with this: HOME PAGE OF MISHA CHERKES

 

Could be drivers issue, try download the latest driver and choose custom install and select clean install option:

Drivers | GeForce

Also there are known issues of RAM compatibility on N56VZ:

http://forum.notebookreview.com/asus/677457-asus-n56vz-16gb-ram-thread.html

You can try open the bottom cover (1 screw) and remove one stick at a time individually and test your games see if the problem goes away.

 

It's an artifact!
Did you change your RAM? Try exchange it with the old one and run a trial
Update your nvidia driver to the latest and also don't forget to update your BIOS because there is one that has 'nvidia' tag.

Try to run it with a fresh copy of windows installation and driver only --> Last thing to do
If all of above has been done, especially the fresh installation, and your problem still there then you should RMA the laptop.

 

Yeah.. non-standard ram configs might cause that, thanks to the forced timing-settings in the bios.

 

test type: DirectX - no errors found.

test type: Cuda (frozen beta) gives a message - Changing video mode to 640x480x16...OK
[7/15/2013 1:10:26 PM] Test started for "GeForce GT 650M"...
Preparing CUDA...Module load error: 300
NO PASS COMPLETED

test type: OpenGL - no errors found.

don't know why the second test didn't work.

 

drivers is up to date, just upgrade bios, but nothing changed. after some ten minutes running Civilization V crash appeared.
for opening cover and removing RAM module I must ask warranty service first.

 

No, I didn't change RAM. Received with 8GB.

 

my RAM modules:

http://img819.imageshack.us/img819/9268/yshm.jpg

 

Looks very much like a GPU problem. Use a Furmark test session and see at what temperature the problem is present if it is. From what I could understand here, should be around 75°C.

 

Blind Men and the Elephant | WordFocus.com

Seriously, though. The crucial ram has lower latency timing than the original ram - except the "drive strength" or tRAS and tRC timing, and possibly a few others. So when the bios forces settings that are similar but slightly below spec, you do see exactly this kind of thing. And yes, it can work fine for weeks before you start to use the computer in a specific way and the error-correction suddenly can't compensate.

Note, ram is perfectly fine. But the forced timings in bios have slightly higher latency, slightly lower drive-strength. Incompatible, because of boneheaded - and persistently boneheaded, after being specifically asked to improve it many times - people at Asus. Their argument, the only one I've ever heard, said that they force ram-timing to ensure no one will overclock the system.

That they won't fix this is akin to a vehicle manufacturer not accepting that putting accordions into all the door hinges is a bad idea. And that they're going to keep the accordions because the designers love polka.

 

test settings and results:

http://img577.imageshack.us/img577/7528/c0b5.png

test ran 1h15min and almost all the time temp was 72°C (from 15min to 75min) somewhere it raised till 74° for a some period, but then decreased again till 72°C. Any suggestions for settings or comments about test??

 

is it a solution to change RAM modules with grater latencies and avoid this problem??

 

Well, that got the GPU out of the way for now. Further you need to start testing the RAM and see how that goes. If RAM is also OK, there might still be a problem with the GPU, but not heat related - most likely it's memory.

I hate these things. I remember having huge problems with Mushkin and AMD in the past, but all the tests were OK. So at some point I tried different RAM and the problem was gone just like magic. But do not buy it just yet, try some first.

 

..yes, that is actually a good point. You will very likely see higher tolerance for the low, "speed tweak", "overclock" type settings they've used on the n56.. and all other asus bioses.. on relatively expensive cl11 ram.

Note that this could all be solved with a very small bios tweak. To make people run around jumping hoops backwards like this, when it's extremely easy to fix is just.. And I'm not just saying that. It's literally about how the entire 16GB RAM thread wouldn't exist if they just hadn't locked two-three settings in the way they've done it. If we had access to the bios functions like you do on any desktop board, it would literally take anyone ten seconds to change it to something that'd work with 16Gb ram of different latencies, different voltage, etc.

I'm really not sure where this comes from either. It's like they've carried over some convention from when sdram could be tweaked down substantially by lowering the drive queue, and following up with the rest. Go from cl3 to cl1 or 1.5, for example. You could massively improve the speed that way. On ddr3 ram, this isn't really how things work any more, and the targets you're going for are never going to, you know, double the amount of instructions potentially shifted, like you could do on sdram. Instead the probability that you're tweaking low and potentially manufacturing ram errors is higher. In return for very, very low improvements in speed or response. Since the ram is designed to run on higher bus-frequencies, and latency is lower in real terms compared to sdram timing anyway, etc.

But it really is just the locked timing that is stopping us from using any ram we want here, with any pre-defined voltage, with such and such profiles, etc. And these settings can be detected by default too - note that - that on any laptop or desktop board made since 2000, those ram-chips can be detected and set properly with no configuration from the user... It really is that bad.

 

Changing RAM modules with new ones solved the problem for the moment. Warranty service said it could be due to RAM modules overheating.

 

*sigh* The... inscrutable mystery persists! For you are -- In the Twilight Zone.. dudiduda dudiduda dudiduda...

 

I also have a n56vz that has problems with games. In games like Arma 3, after a variable period of time (from 5 min to 40 min) Pc stucks (but with no artifacts and got a noise (buzz) coming out from the speakers.
I've tried to install all the lastest asus official drivers and also the latest drivers by nvidia but the problem continues to appear. Do you think it could be a RAM issue also in this case? Wich RAM modules have you changed to? Or as asus warranty changed your RAM?
Thanks for the help!