PDA

View Full Version : G74sx strange GPU behaviour.



IxPanda
01-15-2012, 10:02 PM
Hello Rog, :)

I've been having the strangest experience with my new G74sx. It's a month old right now and it's been doing the same thing for a while now, it seems to be happening more often now. Whenever I play any game, (Skyrim, SWTOR, LoL) anything like that, no matter how graphics intensive I have no problems. however, whenever I go on for example facebook, scrolling through pictures 1 of 2 things happens at random. Either my screen will flicker and the driver will crash for the 560m, or it will actually artifact on me and freeze completely. as for the specs..

actually it happened to me while going to get the specs. x). Here's the driver error:
"display driver NVIDIA windows kernel mode driver, version 290.53 stopped responding and has successully recovered."
and just a fyi, I am using the beta drivers for the 560m right now, but the same thing was happening before I upgraded to them for the skyrim boosts. :)

BIOS Version/Date American Megatrends Inc. G74Sx.203, 23/09/2011
SMBIOS Version 2.6
gpu driver : 290.53

It's strange because I get the driver error, meaning most likely software issues. and then it will artifact up on me, which i was understanding meant it was overheating. Sorry for the novel, any feedback would be greatly appreciated.

fuzon1337
01-15-2012, 10:12 PM
Hi

Have you tried to do a clean install of the nVidia driver when you update them?
When you update the graphic driver you got a option to choose "customize installation" (or something like that) and then you can choose a "clean install" mode when you install the latest driver :)

Optional: Also you could download CCleaner and run the register check to find som junk of files that can cause problems on your notebook (run that check until the program doesn't find any more useless register files)

If it's a overheating issue, blow some compressed air into the fans and see if it's a lot of dust coming out from there.

How is it with latest driver for chipset and WiFi card? :)

IxPanda
01-17-2012, 04:12 AM
Thanks for the quick reply!

I tried a new install, went back to the standard drivers instead of the beta (285.62) for the gpu and ran ccleaner with a bunch of errors and then fixed them all. Unfortunately I'm still getting the black screen / artifacting issues. same error as I previously quoted and the artifacting. Just to give you an example, I was browsing the atheros website (as I'm sure everyone knows really isn't very graphics intensive. x) ) and It actually artifacted from that. I can't imagine it being a heat issue because like I said I can play swtor for hours with no problems. As well the laptop is only a month old, I can't imagine dust being an issue yet. Can you think of anything else?

Thank you.

fostert
01-17-2012, 04:19 AM
Just to check, what are your GPU temps when this happens? Grab GPU-Z or Speedfan to check them. Agreed that it isn't likely this, but lets eliminate it for sure.

PILGRIM
01-17-2012, 05:23 AM
@IxPanda:

How about using a more stable graphics driver (i.e. 285.62 whql)?

IxPanda
01-19-2012, 12:51 AM
@IxPanda:

How about using a more stable graphics driver (i.e. 285.62 whql)?

Thanks for the reply,

Yeah I did revert to that, when I reinstalled the drivers clean I used 285.62 instead, but it did it before I went to the beta as well, very strange..

Another example, I just got home, turned on the laptop and the first page I go to is the rog forum, as i'm loading it it blacks out and tells me the driver crashed. Leads my to believe it isn't a heat issue, but that will be what I do next, check the temperatures and when after a couple times I'll post the results.

I appreciate it guys, Thanks.

dstrakele
01-19-2012, 01:04 AM
Here is an article that discusses how to troubleshoot NVIDIA TDR (timeout detection and recovery) Errors: http://forums.nvidia.com/index.php?showtopic=215256

IxPanda
01-21-2012, 01:09 AM
Here is an article that discusses how to troubleshoot NVIDIA TDR (timeout detection and recovery) Errors: http://forums.nvidia.com/index.php?showtopic=215256

Thank you for the reply,

Just an FYI, I uninstalled the most recent driver and currently am running 268.xx and haven't had any errors for the day, I'll post again in a while if this seems to be a permanent fix or not, but I do thank you for the information.

IxPanda
02-02-2012, 04:40 PM
So i followed the rest of the instructions and did a clean install of the 2.85 drivers and had no problems for a long time, until today when I had taken a confirmation number down in a word pad and then closed the screen, opened it again it artifacted within like 15 seconds... strange.

dstrakele
02-02-2012, 04:51 PM
Leads my to believe it isn't a heat issue, but that will be what I do next, check the temperatures and when after a couple times I'll post the results.

1) So what ARE your CPU and GPU temperatures at idle and when you encounter an artifact problem?

2) When you backrevved your NVIDIA driver, did you boot into Safe Mode and check the "Clean Install" checkbox? If not , you could have NVIDIA driver remnants that are causing issues. You may need to use Driver Sweeper and possibly CCleaner to remove them.

IxPanda
02-02-2012, 05:05 PM
1) So what ARE your CPU and GPU temperatures at idle and when you encounter an artifact problem?

2) When you backrevved your NVIDIA driver, did you boot into Safe Mode and check the "Clean Install" checkbox? If not , you could have NVIDIA driver remnants that are causing issues. You may need to use Driver Sweeper and possibly CCleaner to remove them.

Hmm.. I did not do it through safe mode, I rebooted and it had the driver 268xx on it, the thread nvidia posted said anything below 275.xx was fine, after it was there I went through custom install and selected clean install.. Is this not a clean install? =x .


Also CCleaner has been used, driver sweeper will be next, thank you. I kind of forgot about the temperatures... The problem is now the crashes are less frequent. I'll see if I can catch one though,

Thank you.

dstrakele
02-02-2012, 05:20 PM
You'll probably want to configure logging in your temperature monitoring software. You can get your idle temperatures then run a stress test like Prime95 to see if you can produce artifacts.

That is a clean install, however, I recommend doing it in Safe Mode to disable any running drivers that might interfere with the NVIDIA install.

If I was in your situation, I'd probably attempt to completely remove the NVIDIA drivers, then check to see if there are any NVIDIA remnants in Driver Sweeper, then re-install NVIDIA driver from Safe Mode. If that failed, my next step would be to clean install Windows 7 or restore the Windows System Image of my last known stable instance backup.

Invincible
02-02-2012, 09:20 PM
Here's the driver error:
"display driver NVIDIA windows kernel mode driver, version 290.53 stopped responding and has successully recovered."
and just a fyi, I am using the beta drivers for the 560m right now, but the same thing was happening before I upgraded to them for the skyrim boosts. :)

Yeap i had the same issue , when install this 290.53 drivers but uninstall with uninstall tool and install the 285 WHQL - slove black's and errors
So after that i decided to refuse from such kind of beta

clinthill98
02-24-2012, 11:38 PM
i think i know the problem... i had the same problem back in December.. its the nivdia drivers the 290 ones. i installed them and it randomly started crashing after i noticed that i went back to the factory 268 drivers try it and see and it stopped crashing after that :) i had the same exact problem. get them here http://www.asus.com/Notebooks/Gaming_Powerhouse/G74SX/

IxPanda
08-02-2012, 04:26 AM
Hey again guys. :)

simply to resurrect a fairly old thread. I'll throw down a couple more details for you because just recently I've ran into the problem more and more.. See if this helps anyone narrow down the issue.

- Currently running driver 285.62WHQL , which I have installed via safe mode correctly this time.
- GPU temperature fluctuates from about 45 - 65 When it decides to crap out.
- I've had it happen when reading through an online PDF, watching a youtube video, flipping pictures on facebook... But never In the middle of a game. anything from league legends, to crysis 2.
However, it has happened when the game goes from a loading screen to actual game play. Is it possible theres an issue when my GPU goes from low core speeds and switches to high speeds? Honestly it's the only thing I can think of.

Any further input would be greatly appreciated guys. thank you.

lilysintrigue
08-02-2012, 04:37 PM
Hey again guys. :)

simply to resurrect a fairly old thread. I'll throw down a couple more details for you because just recently I've ran into the problem more and more.. See if this helps anyone narrow down the issue.

- Currently running driver 285.62WHQL , which I have installed via safe mode correctly this time.
- GPU temperature fluctuates from about 45 - 65 When it decides to crap out.
- I've had it happen when reading through an online PDF, watching a youtube video, flipping pictures on facebook... But never In the middle of a game. anything from league legends, to crysis 2.
However, it has happened when the game goes from a loading screen to actual game play. Is it possible theres an issue when my GPU goes from low core speeds and switches to high speeds? Honestly it's the only thing I can think of.

Any further input would be greatly appreciated guys. thank you.

Hi IxPanda :)

I had the same thing happen with my G73sw awhile back, same issues and such... Unfortunately I went through the same steps you did, but ultimately had to send my machine in for numerous RMA's. They had to replace alot of parts, and unfortunately I had to send the machine in for repair 4 times. If you keep trying troubleshooting, and you keep having issues over and over again I would recommend maybe returning your laptop to where you brought it from, and see if they will do an exchange for a new machine. Or if that is not an option, I would recommend you set up an RMA to send in your machine for repair.

Returning the machine would probably be the best option believe me, I survived one of the many RMA horrifying long drawn out repair processes lol .

i hope that you get your machine back fixed without any further issues, and not have to go through numerous RMA's *fingers crossed Best of luck :)

IxPanda
08-03-2012, 02:33 AM
Hi IxPanda :)

I had the same thing happen with my G73sw awhile back, same issues and such... Unfortunately I went through the same steps you did, but ultimately had to send my machine in for numerous RMA's. They had to replace alot of parts, and unfortunately I had to send the machine in for repair 4 times. If you keep trying troubleshooting, and you keep having issues over and over again I would recommend maybe returning your laptop to where you brought it from, and see if they will do an exchange for a new machine. Or if that is not an option, I would recommend you set up an RMA to send in your machine for repair.

Returning the machine would probably be the best option believe me, I survived one of the many RMA horrifying long drawn out repair processes lol .

i hope that you get your machine back fixed without any further issues, and not have to go through numerous RMA's *fingers crossed Best of luck :)

Hello, Thank you for the reply.

The more and more I have this issue the more I'm starting to see that an RMA Seems to be the only thing I can do. :(
I'm just wondering.. There isn't really a way I can guarntee the tech can replicate this error.. any suggestions on how to make it a successful RMA When It's not easily replicated? Also, I put an SSD into this thing.. if I RMA it I'm assuming I'll get the same one back repaired and not a brand new one, so my SSD is safe?

PS I bought it from Amazon.. Not sure if they'd be willing to replace.

Thanks guys.

Woofyhugger
08-03-2012, 04:33 AM
Have you tried turning powermizer off? There is a program that allows you to do this. It was suggested to me when I had constant display driver crashes, and since turning it off, I've not had a single crash.

lilysintrigue
08-03-2012, 12:50 PM
Hello, Thank you for the reply.

The more and more I have this issue the more I'm starting to see that an RMA Seems to be the only thing I can do. :(
I'm just wondering.. There isn't really a way I can guarntee the tech can replicate this error.. any suggestions on how to make it a successful RMA When It's not easily replicated? Also, I put an SSD into this thing.. if I RMA it I'm assuming I'll get the same one back repaired and not a brand new one, so my SSD is safe?

PS I bought it from Amazon.. Not sure if they'd be willing to replace.

Thanks guys.

I'm not sure about the ssd drive, I know that Amazon has a pretty good return policy for faulty products though, I would contact them and see what they say.

If your having the same issues that I was having ie I could play games all I wanted, but when i got on the desktop, or was surfing the net I would get constant kernal driver errors, and your nvidia graphics card has stopped working but recovered successfully.
When this first started, I was only getting them like twice a week, and I was like okay not really a big thing.. But then I started getting them every day, like 15 - 20 a day, tried every driver under the sun, still got them, did clean installs, factory resets everything to no avail. They ended up replacing my motherboard, my mainboard, my hdd twice, and my gpu twice. Basically everything except for the shell lol. If I would have known that I would have had to go through all of that now, I would have returned it immediately, I would contact amazon, let them know what is happening and what you have tried, since you brought it fairly recentely there should be no problem.

That would be the first step I would take, (don't let them push you around :P You paid for a product they sold you, it's defective/ has issues, and they need to stand behind the products they sell to make you the consumer happy. If you have passed their return dates, and too much time has passed since the purchase, unfortunately an RMA is your only option.

I would recommend taking out your ssd and putting the factory hdd back in before you send it in for RMA, because those dullard's at Bizcom will probably be like oh its your hdd (that was the first thing they replaced in mine lol) They won't pay attention to the fact that it's not the same one that came with the laptop, and most likely they will replace it with a normal hdd.

Also, when you call in to the RMA center, let them know you have a G series, and to send you a fedex shipping label, do not get off the phone until you have the label in your inbox, (because calling them back and saying oh I didnt get my label won't work, they like to deny things they have said before lol) I made this mistake and I paid $150 to send my laptop to them the first time via ups w/insurance. Asus does offer shipping labels for their G Series laptops, so don't let them tell you they don't. :)

As for replication of the error you are getting, I like you was weary, but believe me that error will pop up when they are doing the testing! Hope this help's, and if you like you can pm me with other questions!

Sorry this was so long guys :)

IxPanda
08-04-2012, 01:01 PM
Thank you guys so much for all your help.
I'm going to try it out with powermizer disabled.. as that (in my mind) seems to be something that should help.. But of course I really don't know :D.

I really appreciate all the tips for the RMA, I've never had to do one before and if that's what it comes to, I've got a pretty good starting point! :)

if I have any further question's on it I'll be sending you a message, thanks again lilysintrigue, and anyone else who chipped in..

I'll give you guys an update after I've tried it out with powermizer disabled.

Thanks again!

IxPanda
08-05-2012, 11:18 PM
Just a quick follow up with my preliminary findings.

After disabling Powermizer I haven't had a single artifact freeze lock-up OR driver crash... tested with both 285.xx and the beta 304.xx drivers. If anyone is having these issues, I would recommend checking this out. (http://www.overclock.net/t/683583/how-to-adjust-nvidias-powermizer) It talks about performance and low FPS fixes, however It's cleared up all of my crashing errors SO FAR.

PS It gives you an option to keep it running when running only on battery power too, which is nice. :)

Quandth
08-06-2012, 03:14 PM
@IxPanda, How about your temps? Have you noticed higher temps on idle and gaming with Powermizer turned off? I will try this later at home this afternoon, as I have being getting more than usual crash and recovery issues.

Also, do you have to use Powermizer Switch every time you boot like throttlestop? and Do you use Throttlestop along with Powermizer switched off?

IxPanda
08-07-2012, 12:51 AM
@Quandth, I have noticed slightly higher temperatures at idle with it off, Laptop was on while I was at work, Currently running 900/1800/1450 and my temp monitor was 48c for the whole time versus lower 40s. Gaming not so much as it's essentially doing the same thing with powermizer running. You only have to use the powermizer switch and reset it once for it to take effect, it holds until you a) update drivers or b) turn it back on and reset. I do still use Throttlestop with the powermizer turned off, haven't had any conflicting issues. I'd like to hear how it fairs for you Quandth, see if this works for more then just me!

Quandth
08-07-2012, 03:47 PM
I did yesterday the Powermizer switch thing, reboot and started WoW, not even 5 minutes flying around and the laptop freezed, this started to happen when I switched to the 301.42 drivers, I was using before the 295.73 drivers and only the crash and recovery issue happened but not the freezing. I really got tired of this and went straight to the Beta 304.79 and performed a clean install, reboot in safe mode, uninstall drivers, reboot and run driver sweeper and CCleaner, reboot again in safe mode, install new drivers and reboot. After that I ran Powermizer Switch set it to AC Off-Battery ON, and reboot, then I ran MSi AB and changed my clocks to my usual OC settings 915/1830/1625, started WoW and again flying around the driver (304.79) crashed but not freezed, I went and check CPUID to see the clocks, and the driver recovered but to the stock clocks (I didn't know this happened) not the OC ones, I ran MSi AB again and changed clocks to 900/1800/1625, played about one and a half hours and didn’t got any crash or freezing again . Maybe the 30X.XX Drivers don’t like the 900+ clocks, don’t know really, I will run WoW again today for a few hours and see if the freezing or crash happens again, otherwise I will go back to the old drivers, I have being reading several post on other forums about some issues with the 301.42 Drivers and that they were removed from the official Nvidia download page, I went to check it and its true the 296.10 are the newest WHQL ones. I will do more tests today in the afternoon and will post results.

IxPanda
08-08-2012, 04:00 AM
Interesting. Same thing happened to me running the 304.xx drivers as well when I was playing Blacklight Retribution... 915 would freeze, but I can use 900 with no issues at all. Anything over 900 seems to be an issue.

To make sure we're on the same page here Quandth, the crashing I experienced was different then the issues I was having before Powermizer, is that the case with you as well? I'm no longer getting the "nVidia driver xxx has crashed and was successfully recovered." What happened to me at 915 was simply the window closed, I couldn't re open it, but on 900 it was fine!

Quandth
08-08-2012, 09:21 PM
With old drivers at 915 I got the driver crash and recovery issue but was still able to play as soon as it recovered, with the new drivers it will freeze and I have to force shutdown and reboot. When I installed Powermizer Switch and got the crash and recovery issue again I only let it happen once and switched to 900 and everything seems fine, no issues or crashing, although there is an increase in the GPU temp when gaming, about 8C - 9C but nothing out of the normal gaming temps that i have seen other users have reported. So far so good not any noticeable differences in game performance and FPS between 900 and 915.

Edit 08-12-2012: Rolled back to the 285.79 Drivers, better temps on those, with the latest beta drivers the temps went up by 15 to 20C which is not good.

IxPanda
10-04-2012, 01:53 PM
With old drivers at 915 I got the driver crash and recovery issue but was still able to play as soon as it recovered, with the new drivers it will freeze and I have to force shutdown and reboot. When I installed Powermizer Switch and got the crash and recovery issue again I only let it happen once and switched to 900 and everything seems fine, no issues or crashing, although there is an increase in the GPU temp when gaming, about 8C - 9C but nothing out of the normal gaming temps that i have seen other users have reported. So far so good not any noticeable differences in game performance and FPS between 900 and 915.

Edit 08-12-2012: Rolled back to the 285.79 Drivers, better temps on those, with the latest beta drivers the temps went up by 15 to 20C which is not good.

Following up here.

I'm running Windows 8 on my laptop atm, since I've installed it I haven't had a single crash like I explained before, and the only time I've had the "driver has stopped working" error was when I was overclocking and finding the limits.

Interestingly enough, running drivers 306.23 I was able to run 955 Mhz stable instead of 900. I don't know if it's the OS or the new nVidia drivers, but something made a significant impact in performance.