Guild Wars 2

Whenever I try to game with my GPU at stock clocks, my entire PC freezes after only 10-15 seconds in game, in any game. I am locked on the same frame, and all I can hear is this loud buzzing from my speakers. It is probably important to note that I am running on a laptop, more specs below. I have eliminated a few prime suspects…

Memory. Memory tests (RAM) have shown no errors.

Temperature. At the time of crashing, my GPU is only around 60-65 celsius, and CPU is much lower. I also recently replaced the thermal paste between the heatsinks and the die on both my processor and GPU (they each have a separate heatsink and fan).

Virus. Security Essentials and Avast! reports all show clean.

Registry Errors. Using CCleaner, I have cleaned my registry many, many times.

Temporary Files. Also using CCleaner, temp files are non existant.

Now, here are a few more specs.

I am running the most up to date version of Windows 8.

Model: Alienware m15x
GPU: nVidia 460m 1.5GB
CPU: Intel i3 330m 2.13GHz (Dual Core, 2 threads)
Ram: 4.0GB, not sure of the frequencies of each 2.0GB stick, but I highly doubt that would be causing this.
GPU Driver Version: 331.65

I have tried fresh Windows installs, fresh driver installs, fresh game installs, drive defragmentation… nothing has resolved this issue. I have my CPU cores unparked for performance increase. All I run in the background is usually Mumble and Spotify for GW2, but other games, I run nothing in the background and still have the same issue. Underclocking has been the only remedy, but is more treating the symptoms rather than solving the real issue. Lowering my GPU clocks also reduces my framerate by 25-30%.

If anyone could provide any insight into what may be causing this, it would be much appreciated. I can also list any more information I may have missed. Thanks

EDIT: I have also restored BIOS setting to defaults.

Are you sure your games are running on the Nvidia card? Cause if I’m not mistaken Alienware provides 2 GPU’S.

I am absolutely sure I am running on my GPU.

And I’m on a laptop, I don’t see any way of putting in a second GPU, unless you count the integrated gpu on the processor.

And if I wasn’t running on my GPU, why would changing my clocks have an effect?

Yes, I am counting the integrated one.

I have seen the same thing happen before, that the game (or games) use the wrong card. This can be fixed through settings or BIOS.

As for the clocking, no idea.

I’ll take a look into some BIOS settings, but I am positive my GPU is being used.

It’s the only thing I could think of. So I am curious as to how it works out.

I’m not finding anything in the BIOS that can help me

Remove one of your antivirus programs (I would remove the Microsoft one personally). You should never have more than one AV program installed and running at the same time (spyware and malware programs are OK though).

Oh, I just installed Avast! last night to confirm Security Essentials reports of no viruses. I’ll go get rid of it now :P

You could also find the option to use the dedicated gpu in Nividia Optimus. Not sure how that works though, I have an Ati card.

nVidia optimus is technology that optimizes battery life and gpu performance, not a settings interface.

The nVidia control panel is set to performance mode and to use my GPU, and windows is set to high performance mode. My GPU is being properly utilized.

The issue is that when I run at stock clocks, the driver crashes, or my entire PC completely freezes with a loud buzzing sound coming from speakers/headphones. At that point, I have to force shutdown my pc.

Bump

I’m going to try a complete reinstall of windows up to the most recent version. If that doesn’t work, I’ll completely clean my hard drive, and then install windows.

If it’s a hardware issue, I won’t see an improvement.

Windows reinstall had no effect.

Since its obviously a problem with your computer and not related to GW2 i am amazed how the support offers help.

But how about getting in contact with the dell support?
Dont know how old your notebook is or when your problems started.

Hey Ashley, thanks for your response.

I don’t believe my PC has support for optimus. I do not seem to have any options available within my nVidia control panel similar to what is shown in the video.

I am underclocking my GPU with MSIAfterburner. Before that, I was using nVidia Inspector, but after an hour or so, the clocks would be reset to default with nVidia Inspector, resulting in a complete PC freeze again. Afterburner doesn’t have that issue.

I’ll go ahead and create a support ticket. Thank you

Hey crouze,

I have tried getting in touch with Dell support, and they say it’s a faulty GPU probably caused by extreme temperatures. I don’t see how that is possible, since the PC shuts down when the GPU hits 80 celsius, which isn’t enough to cause component damage.

And my notebook is almost 4 years old, and the GPU is almost 2 years old.

Ok, so I did some more looking into it.

I completely uninstalled my graphics driver through Windows Device Manager, then cleaned up using a driver sweeper, then restarted and installed the most up to date nVidia graphics driver.

I booted up HWiNFO64, a program I’ve been using to boost my fans to top speeds to reduce heat, as well as monitor temperatures, as well as clocks and loads on both the GPU and CPU. What I found when crashing was the GPU core load would hit 99%, and the rest of the loads would drop to 0%. At this point, the system would freeze, or the GPU driver would crash and restart, only to crash again within 3 seconds, and repeat until a full system freeze occured.

Screenshot for reference to HWiNFO64

Do you still have warranty?
How about getting a new gpu from dell if they say the gpu is damaged?

Did you ever overclock the gpu?

Because for me it sounds like its not a software issue but the gpu.

Warranty for the PC, probably not. For the GPU, I’m not sure, I’ll look into it.

And for the GPU that I currently have, I did not get it from dell, I got it from a 3rd party source. Unfortunately, that is the only way to get new laptop GPUs.

Ever overclock? Very slightly, maybe another 50 MHz at most into the Core Clock, and 100 MHz at most into the Memory Clock. After seeing no real difference, I bumped the clocks back down to stock. This was quite a while ago, however.

And yes, I am worried it is the card. I would just like to explore all possible software related options before I go out and RMA my card, or buy a new one.

Hi Dreary,

Try Nvidia drivers 314.22, to keep it in short any driver version beyond 314.22 is known to cause instability on Fermi based architecture GPU (that would be Nvidia 400/500 series).

Also regarding overclocking I’d go for reference speeds by nvidia, a full list can be found here: http://en.wikipedia.org/wiki/Comparison_of_Nvidia_graphics_processing_units

Hey Armored,

My stock clocks for the Core clock and the Shader clock match that of what is listed on the wiki page you linked, but my Memory clock is half of what is listed… it’s default it 1250 on my card, but the page lists 2500, which seems quite high.

And I’ll give the 314.22 drivers a try when I can. Any source on that information, though?

Thanks!

EDIT: Still getting the same lock ups and freezing on the 314.22 drivers at stock clocks.

Another thing I have noticed is that the program I have been using to monitor my GPU, HWiNFO64, shows my memory clock at around 635 MHz at load… that might be a sensor error, though. MSI Afterburner shows it at the proper 1250 MHz.

I’ve checked my pc with HWiNFO64 and as I thought it shows the speeds without the “Double Data Rate” effect, If you multiply by x2 you get the correct speeds. Both Afterburner and HWiNFO are correct they just show them differently.

Another thing you could try is removing 1 ram stick and working only on 2GB. With such low ram at best your game will crash to desktop but shouldn’t freeze your pc which would be an improvement. If this doesn’t work remove the installed ram stick and put in the other one to test the other 2GB.

I’ve had cases where I tested ram with memtest86 for 8 hours without an error and in the end it was still the culprit for crashes and freezes.