Guild Wars 2

See the attachmnet for the explanation please…

This doesn’t happen always, but it did happen after a few hours of gameplay, and it happened twice last time i played.
What’s causing this?

When i exit the game, the artefacts are on windows desktop too.
A reboot fixes it, and no other games and applications caused this so far, everything else works good.

So far i only experienced this in GW2 and only after a few hours of gameplay.

What are your computer specs?
List any overclocks with the GPU

I believe it may have to do with any GPU overclocking as well but it may not be the only thing. It could be what Ashley had just suggested. However, I used to use a Radeon HD 5770, armed with the Juniper XT chip. Juniper XT chip did not ever like overclocking and whenever it did have an overclock of anymore than 2% it would just cause problems all across the board, from artifacting to crashing. Funny for an XT chip too as the XT chips are usually the flawless version in a series for the LE, Pro, and XT
I even saw similar problems with many other Radeon HD 5770s and factory overclocked 5770s but they went away whenever the user dropped them to stock settings which I think was 850 and 900.

If you are not overclocking, then its probably a dying GPU Fan. open your system and look at the fan. Is it covered in dust? Can you easily free spin the fan? Does the Fan make a grinding/huming sound when its spinning via power from the GPU?

Those ‘effects’ you are seeing are caused normally by excess heat to your gpu. You can also see them if you overclock to much and stress the Chip in other ways (Voltage, Core speeds…ect). But if you are not overclocking, then im willing to bet 100% its heat and a dying GPU Fan.

Looks and sounds like a overheating and/or dying GPU.

Open up ur case, check over your graphics card, is it full of dust ?
If so, pull it out, take of the plastic shrowd and clean the heatsink and fan.

Befor doign this u could also monitor ur GPU temperatures to find out if overheating is the issue.

If u have added a overclock to it, take it of. if its factory Overclcoked, try downclocking it a little. If it is dying then downclocking could help u get a little mroe life out of it, and if its overheating downclocking (asuming it changes voltages) could help aswell. BUT it shouldnt be overheating, so check for dust and make sure the fans are working.

if it is overheating and it is clean, then you may need to change the TIM, as it could have gone off and is no longer conducting heat well enough for the heatsink and fan to disapate the heat quikly enough.

Thank you for replies!

My system specs (sorry, i should have posted this first):

AMD Phenom II x6 1090t 3.2Ghz
nVidia GeForce 9600 GT 512 MB (it’s factory overclocked edition, XFX Alpha dog edition or something)
8 GB RAM

I don’t know how overclocked is it, but this is the first time it happened.
I didn’t measure the temperature of the graphic card, i’m not even sure there’s a thermometer on it or how to do it (what software, etc.), but i clean my computer with an air compressor once a month, or once every two months, so there shouldn’t be a lot of dust on it.

Guild Wars 2 isn’t crashing at all, the game still works after this appears, but it looks like that. The artefacts remain in the game and after i exit, in Windows 7 untill i reboot my computer.

The fan sounds like a hair dryer on adrenaline mixed with speed. o.O
But that’s how it always sounded, and i’m not sure if it’s the CPU or GPU fan.
And the funny thing is, only in Windows.
In Linux all the fans are quiet and work great, no grinding sounds, no noise. When i start something more demanding i can hear the fan speed up, but in windows, the fans spin like crazy even if the system is idle, and it’s really really loud. That’s one of the reasons why i installed linux in the first place. Something’s not right with Windows i guess.

I checked the Task manager, and everything i could think of, but i couldn’t find any program or process that would cause the processor or GPU to do that. The system is idle and the fans go crazy. When it’s under load (like games and such), the fans are still crazy…

The GPU is old, i know, and yes, it might be dying. I hope not, because i have no money for a new one, but that thought crossed my mind as well.

I could try downlcloacking it, but i’m not sure how to do that.
Also, what is TIM? Thermal paste or something? That’s what i gathered from the context… If so, yeah, that’s a good idea, i never changed the paste on the GPU, but now would be a good time to check that.

One thing i should mention is that Windows behaves weird.
I have no idea why this is happening, i searched the net, but i couldn’t find a fix for it.
Whenever the mouse is idle for about half a minute, and i move it, i get the “Graphics driver has become unresponive and has restarted” error. The screen flickers, Aero is gone, then black, and everything back to normal again.
I managed to add some registry things to prevent the message, and increased the time in which windows thinks the drivers are unresponsive (which should fix it according to their official forums), but the screen still flickers and all, i just don’t get the message telling me that the driver has crashed.
And that’s the second (among many) reasons i got Linux to dual boot.
Alas, Guild Wars 2 doesn’t run well on linux under wine and i can’t play it there, so i need windows (cough make a linux native client hint hint / cough).

Some applications that are currentily using the driver crash because of that, but some just stutter.

Some examples of what happens if i leave the mouse idle for too long:
- Video playback on VLC stops at a frame while the sound continues and i have to restart it.
- Youtube videos go black and need to be reloaded
- Video editing software crashes
-And last, Guild Wars 2 just freezes for a bit, then returns to normal, but never crashes.

Last thing to mention is that the artifacts don’t appear because of the mouse/unresponsive driver thing. They just appear at random, after a few hours, or an hour. There’s no pattern i could reckognize.

There are more examples, but you get the picture.
Sorry for the long post, but i needed to explain the situation.

I really hope my GPU isn’t dying.

I don’t think your GPU is dying, no but it sure is old. Compared to the rest of the system I’d have to wager that it’s a bottleneck at the least. Considering you don’t have money for a new GPU, I suppose that’s a good thing to start saving pennies for haha.
As for the GPU itself,
I recommend looking into reference versions and seeing what the clock speeds are on it and then change your current GPU’s to that (which would be a downclock to reference settings) and see if anything changes aside from maybe a slight dip in frame rate.

Keeping your PC clean is good of you. Monthly maintenance is the best.

As for Linux making the fans quiet, I dunno how that works. I’ve never tried linux before.

Yeah, i think it’s a good idea to start saving money now…
The GPU is definitely a bottleneck, sure, but it serves its purpose. I just want to squeeze some more life out of it before it dies. At least until i have for another one.

And i’m pretty tidy as far as computer hygiene goes. I hate opening up my cousins computer after god knows how long to se dead spiders and all sorts of bugs, webs adn dust in the case. Then I have to clean it of course. Ugh… Once a month cleaning it is.
o.O

Can you recomend some software that i could change the GPU clock with? Or is it a bios thing?

About Linux… Well, it’s free and doesn’t seem to mess up my computer the way windows does lately. It was opposite before. Linux bugged out, and windows, even if it was full of errors kept working, so meh… To each his own. I just use it because it’s preferable to windows now for everyday stuff. I boot into windows when i want to play games, or do some video editing, cryengine 3 editor and such. Programs that don’t work on linux.

Also, for the reference, i attached 2 sound files.
One is the fans on Linux, and the other are the fans on Windows…

When you listen to the windows sound, at about half duration it starts going really crazy…

This is on idle. :/
WTF is wrong with my windows instalation???

Linux fans:
https://www.dropbox.com/s/kupak8wawo6prpt/Linux_fan.mp3

Windows fans (they start going batkitten halfway through the sound so hear to the end):
https://www.dropbox.com/s/c53tqcpn1rbin6n/Windows_fan.mp3

It is a gfx problem, the driver crash you talked about is going to be a corrupt GPU driver install and as such you should go to the OEM website and DL the newest drivers for the card and then uninstall the old one and reinstall the new. If the drivers are not the problem, then it will be over heating with the card, and just because you think it is clean does not mean it is, open it up and check it. It is also a old card, TIM can get hard and lose some of its ability to conduct heat, so it might be worth replacing (its cheap). Also, the fan could be going bad or the card is taking a crap on you. As for checking GPU temps, DL GPU-Z.

It’s one of two things:

A) Your GPU is overheating. If this is the case, you can usually remove the cooler and apply some Arctic Silver – as well as clean the fan (an old toothbrush works well).

B) Your GPU is failing and it’s only going to get worse. You might be able to lower the clock speeds and buy some time – but there’s no way to predict how long it will last.

Ok, thank you all.
I’ll download the latest drivers, do a clean install, see what happens and, I’ll download GPU-Z and check the temperatures.

I can’t do it now, so it’ll have to wait till later today, but i’ll post the results.
Also, replacing the paste wouldn’t hurt, so i’ll do that too at some point.

So one program you can use to help with the fans and also for overclocking and downclocking is EVGA PrecisionX, it also monitors stuff aswell.

GPU-Z is the best program to use for pure monitoring though so yes do get that.

As for ur windows issue. I always find it easier to do a fresh install if i think there is a corrupt driver or file within windows. However thats just me, u can always try a system restor to a point when u remember the issue not being present.

TIM = Thermal Interface material. So yea thermal paste.

I have 2 old 8800GTX reference models stored in my loft and i remember how loud those were once they started ramping up the fan speeds, so i dont think anything is wrong with ur fan just by going by the noise. The XFX model as far as i can tell uses more or less a reference design hence why i compare it against my old 8800 GTX’s.

The fact u dont hear the fans ramp up in Linux might actually be a bad thing, if anything as it means ether ur GPU isnt being properly used, or its letting it get to hotwithout uping the fan speed.

100% and issue with video memory failing. Check the following:

1. Fan on video card (overheating).
2. Overclocking of GPU (overheating).
3. Seating of video card in PCI-E slot (just reseat it to be sure).

If all of those check out, your video card COULD be going bad (possibly a bad solder joint on the board), but doing a clean reinstall of the drivers couldn’t hurt.

If this happens intermittently, then I’d lean toward a bad solder joint that is failing.

Every time I’ve seen that kind of corruption it has been bad video memory.

If after all else fails and u need to get a new card, there is 1 possible way to MAYBE save the card. Its a risk, but..u know if its dying/dead then ur not loosing anything if what i sugest goes wrong.

Its called baking. Even if u do it right it may not work, so just putting that out there.

You bassicaly stick ur graphics card in the oven. Sounds mad ..yep ..and it is. BUT do it right and it can bassicaly resolder everything on the baord at the same time, thus fixing any potential bad solder joints.

What can go wrong ..well u bake it to long or to hot and its fried…erm roasted.
The chips and caps could fall off and can happen if u make it to hot or leave it in to long.

The process, well the ones iv seen bassicaly strip the card of the heatsinks, shroud, and fans, stick it on a foil lined oven tray, suport it with somthing like rolled up balls of foil, and place it in. Have the oven preheated to around 195c/385f befor hand and leave it on for 8-10 minutes. it should start smelling of solder. After that let it cool slowly.
Once done, reassemble it and stick in ur rig and see if it worked.

There are alot of vids on youtube showing other people doing it and also forum posts on places like overclock.net.

Again only do this is all else fails and ur about to buy a new graphics card.

SolarNova.1052:

If after all else fails and u need to get a new card, there is 1 possible way to MAYBE save the card. Its a risk, but..u know if its dying/dead then ur not loosing anything if what i sugest goes wrong.

Its called baking. Even if u do it right it may not work, so just putting that out there.

You bassicaly stick ur graphics card in the oven. Sounds mad ..yep ..and it is. BUT do it right and it can bassicaly resolder everything on the baord at the same time, thus fixing any potential bad solder joints.

What can go wrong ..well u bake it to long or to hot and its fried…erm roasted.
The chips and caps could fall off and can happen if u make it to hot or leave it in to long.

The process, well the ones iv seen bassicaly strip the card of the heatsinks, shroud, and fans, stick it on a foil lined oven tray, suport it with somthing like rolled up balls of foil, and place it in. Have the oven preheated to around 200c/385f befor hand and leave it on for 8-10 minutes. it should start smelling of solder. After that let it cool slowly.
Once done, reassemble it and stick in ur rig and see if it worked.

There are alot of vids on youtube showing other people doing it and also forum posts on places like overclock.net.

Again only do this is all else fails and ur about to buy a new graphics card.

Its true, But you dont JUST put the card in the oven.

You need to take off anything that is detachable that is Plastic (your HSF, for example).
Then you MUST cover the card in TinFoil and cut out squares where the RAM and GPU DIE, are. You must absolutely cover any exposed Capacitors or they will Explode under the intense heat. (TinFoil acts as a heatshield, reflecting the Heat away from sensitive components on the PCB. )

then, preheat your oven to 375degrees, and bake on an open rack for 15-20minutes.

I have done this on many PC Components, Including Laptop Motherboards that have a ‘No Video’ issue, and the success rate is about 92%.

Most the vids, writen tutorials, and forum posts iv seen havnt bothered insulating anything specific on the cards. I guess it wouldnt hurt, but i imagin it would be a bit of a PITA to do so. And most do ~10 mins. But im sure there are other methods/time periods that have worked for other people. its a last resort so what ever floats ur boat i guess lol.

Ok, i must say i’ve never hear of baking a graphic card. o.O
That’s a new one. :P
Thanks for the tip, but i’m not gonna stick it in the oven just yet, because, as you said, if anything goes wrong, i could fry it for good, and i don’t have enough saved for a new one.

But i downloaded GPU-Z as suggested, played Guild Wars 2 for about an hour or two, and set it to tell me max readings (not current ones, but what the maximum was at any point).

The results are in the attachment.
Is this too high?
And this wasn’t even with any excessive gameplay, i did 2 story quests and ran around the map doing vistas. There were no players around as it’s the time every normal person sleeps so… Yeah… :P

The fan was spinning loud, but not as loud as when there’s an event and a bunch of players show up, and the whole screen lights up with skill effects, so i suspect the GFX gets hotter then.

Tomorrow i’m getting some thermal paste, and i’ll apply it. Can’t hurt…

So… Is heat the problem here?

EDIT: I forgot to mention, i didn’t reinstall drivers yet, haven’t gotten arount to it.
When i do a “clean” install, what does that mean? Do i just uninstall drivers and install new ones or what?

Temps are high for normal load, shut down temp for the 9xxx I think was around 100C, and normal load temps were in the 60C range.

Clean install would mean uninstalling the current drivers and settings, rebooting and reinstalling the new drivers.

Here are some steps for you to take.

1. Shut down computer and clean it the best you can. Take off the little fan shroud on top of your video card and clean out the dust with an old tooth brush as others have stated. Be gentle.

2. Boot up and uninstall your old nvidia drivers from the control panel add/remove programs. Reboot and then delete any nVidia folders you find within program files if there is any present.

3. Go here http://www.geforce.com/drivers and choose the geforce 9 series and then your operating system, with 8GB of RAM I assume you have 64 bit. Example: Windows 7 64-bit. Download and install the latest WHQL driver from nVidia which is currently 331.65 – WHQL. It will probably ask you to reboot after install.

Hope this helps and to further address your high heat problems make sure you clean everything up in your case and under your gpu fan shroud.

Also if you are interested in downclocking your card to it’s stock clocks I would recommend MSI Afterburner this program should work.

Stock clocks can be found here: http://www.geforce.com/hardware/desktop-gpus/geforce-9600-gt/specifications

As for the defaut voltage all I could find on it is this from techpowerup.com “Default vGPU voltage for the GeForce 9600GT is 1.10 V at idle”

if cleaning and drivers don’t find your problems then I would try the msi afterburner and setting clocks to their reference clocks.

If you need to know something do not hesitate to ask.

86c at full load with almost max fan speed does see ma tad high.

Though ill be honest, i cant remember the average temp people used ot get back when those cards were mainstream. Nvidia’s thermal shut of is at 105c, but the higher ur temp the less stable thigns can get. It might be showing its age and thus need lower temps now to run.

Change the TIM, re-install drivers, and see what happens.
If that fails try downclocking it.
if that fails then last resort before calling it ‘dying’ would be to format ur whole system, re-install windows, re-install drivers, and re-install GW2 and see if it still does it.

Another thing u can do is find another rig and try the card in there just to see if its deffinatly the card.

Veprovina.4876:

Ok, i must say i’ve never hear of baking a graphic card. o.O
That’s a new one. :P
Thanks for the tip, but i’m not gonna stick it in the oven just yet, because, as you said, if anything goes wrong, i could fry it for good, and i don’t have enough saved for a new one.

But i downloaded GPU-Z as suggested, played Guild Wars 2 for about an hour or two, and set it to tell me max readings (not current ones, but what the maximum was at any point).

The results are in the attachment.
Is this too high?
And this wasn’t even with any excessive gameplay, i did 2 story quests and ran around the map doing vistas. There were no players around as it’s the time every normal person sleeps so… Yeah… :P

The fan was spinning loud, but not as loud as when there’s an event and a bunch of players show up, and the whole screen lights up with skill effects, so i suspect the GFX gets hotter then.

Tomorrow i’m getting some thermal paste, and i’ll apply it. Can’t hurt…

So… Is heat the problem here?

EDIT: I forgot to mention, i didn’t reinstall drivers yet, haven’t gotten arount to it.
When i do a “clean” install, what does that mean? Do i just uninstall drivers and install new ones or what?

Your card is over clocked, per the standard config from Nvidia.com;

http://www.geforce.com/hardware/desktop-gpus/geforce-9600-gt/specifications

Yours Standard
GPU Speed 740 650
Memory Speed 1000 900
Shader Clock 1850 1650

So I dont know if your card comes like that out of the box (Cards BIOS Setting from the Vendor). But if you are still having heating issues, your Fan checks out, you have no Gaps between the Nvidia Chip and the Heatsink. then You might wanna run reference Clock speeds.

sirsquishy.8531:

So I dont know if your card comes like that out of the box (Cards BIOS Setting from the Vendor). But if you are still having heating issues, your Fan checks out, you have no Gaps between the Nvidia Chip and the Heatsink. then You might wanna run reference Clock speeds.

It came like that, i never overclocked it myself…
I finally got that thermal paste, so tomorrow i’m doing the cleaning, and i’ll disassemble the heatsink and apply a new thermal paste layer on.

If that doesn’t help a lot, i’ll use some program to downclock it to the default settings.
I’ll post the results when i finish and test it.

Thank you all for all your help!
Keep your fingers crossed!

EDIT:
Also, i did a clean install of the new drivers. And i really mean clean. I removed all the Nvidia leftovers manually, all the folders and registry keys, and just to make i went for the overkill. When i installed the new drivers, i checked the “clean installation” box, because, nvidia apparently can do that now, and that’s supposed to remove all the leftovers, but i don’t trust that much… :P

I’m happy to report that (for now, i don’t wanna jinx it), the annoying, “Nvidia driver has stopped working” bug is gone!
Finally!

As for clean install and uninstalls, both NV and AMD are far better, in the good days, they use to leave crap everywhere, but that is long gone and the uninstalls are pretty good, while it is best to do clean installs, more so when having problems, but I have been doing updates right over the old drivers for well over a year now on many systems without a single problem.

Also, replacing the TIM, be sure to clean all the old TIM off first, and when putting on new TIM to not use allot, people tend to use to much which can cause higher temps. You want to use a dot or line in the middle of the GPU die which just enough so when putting the heatsink back on it spreads out just enough to cover. Depending on the TIM you got, the mfg’s site will have a how to of putting it on with pictures and all in most cases.

Yeah, the drivers do seem better at cleaning themselves, but idk, call me paranoid. :P
I just had to do it.
I would have just updated it, but last time that didn’t work and the bug was still there, still driver stopped responding etc…

So why not. The drivers did leave a lot of crap behind upon uninstall. Maybe the new installer cleans that first, then installs new ones, but i can’t be sure. It says “clean install” but hey… Paranoid. :P
Also, probably some of the registry keys were causing the issue, or maybe the driver version.

Don’t know for sure, but whatever it was, it’s gone now.

Now i just need to revitalize my GFX card with new paste, and clean it.
Hopefully i’ll get good results in the end.

I already got some tutorials bookmarked for applying thermal paste, so i’ll check them out before i do anything.

Well, i made it! Cleaned the fan, heatsink, the rest of the thermal paste, applied a new coat, and the computer booted up. Yay i didn’t mess it up! :P
I didn’t go overboard with the new layer or paste, i applied a medium thin layer. I read that if you put too much that it’s not good.

I ran the tests on idle and under load (about an hour of Lion’s Arch, and zerg), and the results are in the attachment.

I’m quite happy. Not as low as it could be, but it’s better than before. I’ll try downclocking it later on, that should prolong it’s life some more, and when i downclock it the temperatures should drop further too.

I haven’t noticed any artifacts since that last time, and with the new drivers, all the driver crash problems are gone! So i think the heat and the drivers were causing the problems, as there was virtually no thermal paste left on the GPU, and what little there was, didn’t look all that good. Probably worn with age.

So i think this is it. I can’t say for sure until i play for hours like the last time this happened, but it’s certainly better than before and the artifacts shouldn’t return.

Thank you all for your help!

Glad you got it resolved! Temps look good to me.

[Sigh] It’s time for me to clean the dust out of the fan on my graphics card because it’s gotten noisy. If I wasn’t working on another PC today I would do it – OK, well…actually…I’m just too lazy to do it right now.

Hahah, well it’s amazing what a bit of thermal paste and an air compressor can do. Mine was really full of dust, despite the fact that i clean it once a month. No wonder computers get slow with all that crap piling up in there. Add to that the age of the hardware, buggy drivers and yeah… Artifacts on screen.

I actually kinda enjoyed tinkering with the hardware. It was fun, and i’m really glad i got rid of that “Nvidia driver has stopped working” bug. It was driving me crazy.

Again, thank you all for your help! I really appreciate it!

I should check out that old Radeon HD 2900 XT I have and see if I can separate the heatsink from the PCB. So many screws…

Avelos.6798:

I should check out that old Radeon HD 2900 XT I have and see if I can separate the heatsink from the PCB. So many screws…

I only had 4 to unscrew to remove the heatsink from the chip.
I couldn’t get the protective top off from the heatsink though. I couldn’t find such a small screwdriver… If i could i would probably clean it better, but meh, if it works, it’s fine, and i’m happy with the temperatures.

My 2900XT has a backplate too so I have to basically take off everything.

Avelos.6798:

My 2900XT has a backplate too so I have to basically take off everything.

Ah. Then yeah… Way more screws than mine.
Be sure to pack a variety of screwdrivers, some of them (at least on my GPU) were pretty small…

I have a tiny screwdriver which i use for my glasses and some guitar parts.
The screws on the heatsink plate were even smaller. It was too big for them lol. :P