A curious GW crashing problem

So here's a little problem I've encountered with Guild Wars and haven't been able to fix for several months now. I have ran out of ideas by now and I hope someone will be able to help me out.

Around March I custom built a PC with the following specs:

- Gigabyte MA790X-DS4 motherboard
- AMD Athlon64 X2 running on 3,4 Ghz
- 2 Gig Kingston Dual Channel RAM
- Gigabyte GeForce 8800GT
- Corsair VX 450W PSU

Until this summer Guild Wars ran just fine on it, I had no problems whatsoever. One day it just started shutting off my monitor at a random interval being in the game (I knew the PC still ran, I had the music in the background). After setting up Service Pack 3 it got even worse, with the game shutting the monitor down, emiting a beep (only sometimes) and rebooting the PC, after which the keyboard would lose power like there was no power feed to it.
I tried every single solution I could think of: install the newest drivers possibile under SP3, running various diagnosis checks which reported nothing wrong, I removed SP3 afterwards, reinstalled first the drivers that I had gotten with the hardware, then tried installing the newest drivers after that proved to be inefficient. I tried lowering it down to one core, with no success. The PC is currently up to date with every Windows update.

The curiousity of it all, aside the fact it just happened out of the blue, is that it does it only to Guild Wars. I can run Unreal Tournament 3 or Fallout 3 with maximum details and the performance didn't suffer one bit at all or manifest any of the similar problems that Guild Wars does.
I had also submitted a NCSoft support ticket but the only answer I got after hours of diagnostic was "take your PC to a PC shop".

Thank you in advance for any suggestions.

2 quick thoughts:

your CPU is overclocked isn't it? That can cause system instability and GW is a sensitive game, esp where heat is concerned.
Also a 450watt PSU is lower then is usually reccomended with a 8800GT, that could be causing trouble in the same way as the CPU.

Those are just 2 ideas that popped into my head while I was reading your post. If I think of anything else I'll post again later.

The CPU is not overclocked, I don't have the time, money or enthusiasm to overclock CPUs. I have also assembled the system in a Raidmax Sagitta chasis which comes with 3 fans on it so temperature is seldom a problem. (I had to get a bit extreme on that, the summer temps used to kill my older PC which would run only 1 fan)
As for the PSU, when I was buying it, I happened to buy both the PSU and the card in the same store and I was told it will work fine whatever I decided to run.
I thought of opening the case and rearranging the cabling how I could, but the point that it happens only to GW and not other games kind of dissuaded me from it. I might try it though, just to see what happens.

The obvious next step would be to try a different 8800GT and see if it's the card itself, but that's not that easy (or cheap) to do.
GW does utilize some DirectX routines that some of those other games don't, so it could be drivers or even directx.
Or, it could be the power supply......

Have fun.

I would lean more toward the power supply. You have to keep in mind that your power supply is sending power to every component in your rig. Obviously it's not a driver or update issue since you have everything up to date.

You might consider replacing your PSU. Now that it's around christmas time you can get some pretty good deals. Last christmas when I built the rig I have now, I was able to get a 550w thermaltake PSU for around $20 after rebates. At least that would check a PSU problem and fairly cheap to test.

I dismissed the PSU as a problem because the same thing does not occour when running games that are more requiring: Unreal Tournament 3 on maximal details possibile does not cause anything like it no matter how much it runs, Fallout3 and Neverwinter Nights 2 are running also flawlessly. Also, where I live currently quality PSUs are not too cheap, even with the Christmass season coming up.

I might have fixed the problem today by limiting extensions with nVidia Control Center in Advanced mode. It did not crash for 2 hours but I guess I'll be only able to tell next weekend when I come home from college again and try it after a new booting.

You could try forcing GW to run in one core, but it doesn't sound like the likely solution in this case.
But, to try it, start up GW. While GW is running, hit ctrl-alt-del and start Task Manager. Under the "processes" tab, find "gw.exe". Right click on it and go to "affinity". Uncheck one of the cpus and exit Task Manager. If GW runs without crashing, you can use ForceCore to make GW run in one core every time.

Tried that too before, but it didn't stop it from crashing as usual.

I can't try anything else till I get back home for the weekend, but my hope is that I fixed it last weekend by limiting extensions with the graphic drivers.

Well seems I'll be biting my tongue on that one.
It ran less than what it ran last weekend, then it shut down the PC and it didn't want to turn on for about 10 mins. I'll have to meddle with cables tomorrow I guess...

In the meantime I have reinstalled DirectX 9.0c in hopes that'll do it. Otherwise I'll have a hard time convincing my parents this thing needs a new PSU.

Any good quality power supply will have overload protectors which will shut the unit down if they detect an overload or overheat condition. To prevent the overload from causing permanent damage, these circuits will either remain ON for a period of time and/or require that the AC main power be disconnected and reconnected.
At any rate, it does sound like the PS is going into shutdown.

Open up your computer and blow any dust and dirt out of any and all fans and heatsinks you can find - including those in the power supply.
Caution - if you see dust/dirt in the power supply, you would need to remove the power supply and open it up to properly clean it. Be sure to disconnect the AC power before you take it apart.

While the side of the computer is open, power it up and make sure that all the fans are running.

If all the fans are running and the heatsinks/fans are clean(-ish), the power supply or some other part may have simply gone bad. The easiest way to find out what part is bad is by direct substitution, but, of course, that's a bit difficult for someone who doesn't have a bunch of spare parts kicking around.

The only help I can give you with your parents is that you can tell them that the retired owner/operator of a computer business thinks that something is definitely wrong, and it will probably only get worse.

I cleaned the whole thing about 2 weeks ago. The drawback with this PSU is that the fan points into the case, not out of it, so it's a bit hard to see.
The fans were my 1st suspicion, I checked them quite easy since I have a plexiglass side, although I had to stick a small mirror under the graphics card because I still value my fingers :P The fans work.
I'll have to pry open the PSU I guess.

As far as the spare parts go, I have em, just all of them are old setups and no PSUs remaining around.

I believe either your gpu or cpu is running hot. There are a number of posts in these forums that describe similar problems and all seem to have a common theme. They all have an 8800 series or 9XXX series Nvidia graphics card. Before I changed to an ATI video card, I needed to add a Zalman Cooler to my video card to keep my machine from beeping or shutting down.

You can get an applet from the Nvidia web site that monitors the temperatures. It may even be on the disk that came with the card or mobo.

You can also raise the warning temp that sets off the beep or shuts you computer down, in your BIOS. I wouldn't suggest doing that but it is possible to do.


"...then it shut down the PC and it didn't want to turn on for about 10 mins."


Yeah, it won't turn on until it cools down.

Well a Zalman cooler is already on the graphics card, it came with the card.

The nVidia temperature monitoring tool is meant only for the motherboards running an nForce chipset, which is not what my motherboard has.



This is my temp reading after a 3 day run 24/7 on maximal load. (I'm not sure what Temp3 reads, the 3 chassis fans are not shown either because they don't have any way to read out).

Right, bumping a year old thread, but after the modifications I've done over this year, the issue still seems to persist.

My PC specs are now these:
- Gigabyte MA790X-DS4 motherboard
- AMD Athlon64 X2 running on 3,4 Ghz
- 4 Gig Corsair Dual Channel RAM
- Palit GeForce GTX275
- Corsair TX 650W PSU

I swapped out the PSU, the RAM and the Graphics Card thinking each in turn was the problem. One of my friends was even in contact with NCSoft and has been told that it was a common problem brought by a PSU/Graphics card combo. Doesn't seem to be the case here. I reinstalled Windows about 3 times and ran GW on Linux (that too crashing after a bit of running).
The graphics card is running on a decent 45°C while running GW and the CPU is running at 41°C. The one thing I found is no Windows logs, just treating the reboot as a badly done shutdown.
Any ideas what else it might be? Because I'm out of ideas.

I re-read your thread and I didn't see anything about re-installing GW? It's a simple and easy solution and if your gw.dat file is corrupted it will fix it. In any case, nothing to lose. BTW, your temps are excellent load temps.

have you tried updating your Mobo's BIOS? looks like the current BIOS is from Aug. 12th 2009.
http://www.gigabyte.com.tw/Support/M...2695#anchor_os

(Make sure that is correct MOBO/BIOS before updating/flashing.)

depending on OS, there may also be mobo driver updates as well.

/edit
you might also want to check the memory compatibility list. The last gigabyte board I went up against was using memory that wasn't on their supported list, however it worked, but the BIOS had it over-volted to 2.2v and the timings were all out of wack. had to set the RAM to factory settings/timings and work from there.

BIOS flashed, both the memory from Kingston and Corsair run fine (had em running thru some memtests originally to check for problems), all drivers are fully updated.
And reinstalling Guild Wars came with reinstalling Windows sorta, I wipe the GW dir every time I reinstall Windows so I keep my skills saved, but do a clean reinstall. What still baffles me to this day is that GW worked fine for the 1st 4 months I had the PC (old specs) and then out of the blue it screwed up.

@ElderIII: I was amazed myself that the Palit GTX275 showed 35°C on pure idle, and only 55°C when running Arkham Asylum.

have/could you try a different monitor? poking and hoping. cables..etc..etc..etc..

Yep. Tried. The monitor I'm using now is not the same I had at the start of the problem either, I used a CRT back when I first asked for help here and now I'm using an LCD one. And I swapped out 3 monitors out looking for the problem.

the only things in common between your old setup and the new one are the CPU and motherboard. so the problem is caused by one or the two. since you are running an overclock on the CPU (the highest stock clock on an athlon 64 x2 is 3.2ghz), the first thing i would do is to set it back to stock speeds and see if the problem persists. if it does, then maybe your motherboard is to be blamed.