Sign-in / Sign-up
Your question

PC crashes strange while playing games

Tags:
  • Nvidia
  • Games
  • Crash
  • Graphics
  • Product
Last response: in Graphics Cards
June 18, 2012 8:11:14 AM

PC crashing strange while playing games?
So here's the deal.My desktop starting crashing very hard randomly while i play games like Supreme Commander 2,BF3,COH(tales of valor).The strange thing is that when it crashes different artifacts appear everywhere and i have to restart the PC,but sometimes it crashes with artifacts,but the game closes and i can start other programs like Mozilla,other games etc. but they crash right away and again i have to restart.But windows programs like My Computer,Recycle bin don't crash and work properly.First i thought it was overheating so i got two Case fans Which blow right at the Video card.It worked for sometime but then it again crashed(same way).The one thing that suprises me is that whatever solution i find the ''CRASH" finds a way to counter it and happens again(i know it sounds stupid).I have the case open from the right side.Lasted a little longer but again crash.And now i have no idea what to do.I used MSI Afterburner and put fan speed to 100%(same).I tried other stuff too but still nothing. PLS help.

System Specs:
Windows 7 64-bit
Processor:AMD Phenom II X4 850 3.3Ghz
Video card: Palit Nvidia GeForce GTS 450 2.7GB
RAM 4 GB
2 extra case fans(TURBOX TX-FL80 80MM 4LED 2500RPM)and(Cooler master 120MM)
650w Power Supply(previous was 400W)

PS:When the pc crashes sometimes there is a error message and sometimes there isnt,plus there is a strange sound like (BZZZZSS).Temps are normal 60C max 70C(while playing)35-40C idle.Could the processor be overheating or something else i have no idea.

Thanks.

More about : crashes strange playing games

June 18, 2012 8:39:25 AM

Have you tried monitoring your temperatures? Trying to increase cooling while you don't even know it's overheating which causes the crashes isn't very smart.
June 18, 2012 8:42:33 AM

Implement a remedy: buy a new PC.
Related resources
June 18, 2012 8:45:22 AM

From what i can see you upgraded your power supply recently and that is when the problems started happening. A flaky power supply can cause a large number of head aches ... I wouldn't use the 400 watt you had previously (seems a bit low). Try returning or RMA the power supply you just bought and try to get another. Also if your tech savvy enough you can use a pocket multimeter to test the voltages on the power supply cables.

Graphics cards don't like being under-volted. Processors usually wont function if they are under-volted and or blow up if they are over-volted. But more than likely it just isn't putting enough voltage to the graphics card. I've seen this happen in many of my older builds.

The reason why your my computer/ recycle bin / desktop worked was because the graphics card has a 2d mode where it under-clocks the entire card when it doesn't detect a game running which has the effect of lowering the power required from the power supply.
a b Î Nvidia
June 18, 2012 8:56:48 AM

Generally when people only say the wattage of the PSU its cause they dont know what it is or its absolute garbage. Id start there.
a b Î Nvidia
June 18, 2012 9:00:29 AM

What power supply? run furmark for atleast 5 minutes and watch it, make sure the temps aren't going above 80c ish.

If it doesn't crash from that, run prime95 for atleast 5 minutes, monitor temps, make sure it doesn't get above 60c ish.
June 18, 2012 9:39:32 AM

I suspect the GDDR is overheating, try underclocking ( by say 25%) the memory with GPUTool. If this allows you play games without crashing you know it is the culprit.

Most GFX cards do not have sensors on the GDDR (mine does but its kinda special). So it is often hard to diagnose. Does your H/S fan make contact with the GDDR chips? If so you may have a TIM pad missing.

If not, it is more likely that some of the RAM is de-soldering slightly. The oven bake trick is the only known "permanent" fix (obviously it is not forever permanent, but may last a good length of time).

Underclocking your graphics RAM will usually not result in huge performance decreases, depending on res and details (the higher the more impact).
June 18, 2012 10:02:43 AM

What are the models of both PSUs? It seems the problem started happening right after you got the new one.
June 18, 2012 10:53:14 AM

^^ he never said when it started happening or any sort of timeline for the PSU change.

He says that taking active measures to counter heat build-up help but don't cure. It seems definately a thermal issue...
June 18, 2012 11:41:53 AM

Turbo-X (Solytech) 650W Value Series:that's the power supply.
And i do think its overheating cause it when i put the new fans it worked properly.
But im not sure what is overheating.
Plus the problem stayed before i bought the new power supply.
a b Î Nvidia
June 18, 2012 11:49:34 AM

Well, follow my advice and see if the problem is indeed overheating.

And I can't find anything at all on that power supply...
June 18, 2012 4:58:07 PM

kemperkipie said:
Have you tried monitoring your temperatures? Trying to increase cooling while you don't even know it's overheating which causes the crashes isn't very smart.

maybe
June 18, 2012 5:42:08 PM

I ran furmark while my card was underclocked by approx. 25%.
Stayed at 60C
I also ran prime95 for about 25 mins at max and the CPU temp was 57C max 58C
June 18, 2012 5:53:27 PM

Was it the Memory or GPU that you underclocked?

Was it stable?

I seriously think it will be the GDDR if anything as both CPU and GPU temps are well low.

Make sure you underclock the memory not the GPU. Some cards have multiple profiles (HI-HI, HI-LO LO LO-LO) etc. Make sure you underclock the memory on all of the 3D profiles. SAVE BETWEEN EACH PROFILE, you can't change them all then save. So set clock, save, change profile, set clock, save etc.
June 19, 2012 8:37:08 AM

I underclocked the GPU but i will return it to stock settings.It worked for while but it crashed again.
By the way how do i underclock the GDDR.
with a program or what?:) 
a b Î Nvidia
June 19, 2012 8:40:21 AM

Hes just referring to the memory clock rate instead of the core clock rate.

I would run memtest until it completes (usually takes around 3-5 hours so do it while sleeping) just to rule out any memory related issues.
June 19, 2012 5:00:06 PM

I ran memtest and it didnt find any erros.Plus i used MSI afterburner to underclock it but still nothing. :(  Any more ideas that might help? :??: 
June 19, 2012 6:32:04 PM

afterburner does not let you go really low with the underclocks, normally only like 50mhz or something (this may be model specific).

Clock it down as low as you can with GPU tool pls. Verify that it is clocked down in ALL 3D modes GPU tool lets me clock much lower than afterburner for my card...

Re-seat the graphics card too, just in case it is a dodgy connction.
a b Î Nvidia
June 19, 2012 6:39:10 PM

Hm, yeah afterburner lets me get down to 400 core speed from the reference 900~, so probably card specific.

If you know anyone with a capable power supply you could try that.

Have you tried uninstalling current drivers, restart into safemode run driver sweeper, clean out amd display drivers (NOT CHIPSET just display), restart out of safe mode, install most current drivers?
June 19, 2012 6:51:13 PM

mouse24 said:
Hm, yeah afterburner lets me get down to 400 core speed from the reference 900~, so probably card specific.

If you know anyone with a capable power supply you could try that.

Have you tried uninstalling current drivers, restart into safemode run driver sweeper, clean out amd display drivers (NOT CHIPSET just display), restart out of safe mode, install most current drivers?

im with nvidia not amd but i understand :) 
June 21, 2012 6:33:27 AM

i did both things and kinda worked for a while.It looks like that on small maps with 24 players works,but when i got to a map with 32 players it crashes at some point.
June 21, 2012 9:41:17 AM

Well at this point I would say,

FInd a spare Graphics card around, an old one perhaps, or a friends. It is just to test for sure that it is what is causing your fault.

Play a game with a different card in for a while, if there is no crash (I suspect there won't be) then we can say quite safely its the graphics card giving you problems.

If this is the case, I offer you two choices.. Easy option, buy a new card

Harder Option, strip the graphics card down to the PCB (take off H./S Fan and any plastic mountings etc.) Pre-heat an oven to 160C, for 20 mins.... Let it cool slowly for at least 30mins. Re-assemble and cross your fingers.

Here is a vid: http://www.youtube.com/watch?v=UkDSEbHV8U0

This MIGHT work and save you some money. If it doesn't you will need to buy a new card anyway so what the hell?
June 21, 2012 5:59:42 PM

americanbrian said:
Well at this point I would say,

FInd a spare Graphics card around, an old one perhaps, or a friends. It is just to test for sure that it is what is causing your fault.

Play a game with a different card in for a while, if there is no crash (I suspect there won't be) then we can say quite safely its the graphics card giving you problems.

If this is the case, I offer you two choices.. Easy option, buy a new card

Harder Option, strip the graphics card down to the PCB (take off H./S Fan and any plastic mountings etc.) Pre-heat an oven to 160C, for 20 mins.... Let it cool slowly for at least 30mins. Re-assemble and cross your fingers.

Here is a vid: http://www.youtube.com/watch?v=UkDSEbHV8U0

This MIGHT work and save you some money. If it doesn't you will need to buy a new card anyway so what the hell?

ok..... but arent there any other ways cause right now i dont have money for a new card.I will put a friend's card to check but still arent here any other ways? :) 
a b Î Nvidia
June 21, 2012 6:06:56 PM

You could always try buying some heatsinks for the VRM chips, but this is a last resort (I would do it before trying americanbrians suggestion though) and should only be done once 100% certain that, that is actually the issue and is something not power related.
June 21, 2012 7:17:48 PM

Well i think it is the GTS because CPU isn't overheating,i have enough power,i dont like cooking my pc parts.Also everything i do to cool the video card works but doesnt cure.So i really think that teh card is overheating.Can you give me some links for these heatsinks you are talking about.Please.
June 22, 2012 6:19:14 AM

It does say ''ATI & NVIDIA Compatible'' so doesnt that mean all ati and nvidia cards? :??:  Also what are these?
a b Î Nvidia
June 22, 2012 6:32:43 AM

They go on all your little fiddly bits on your gpu that might not have heatsinks, like VRM, ram modules, etc.

I think you can check your vrm temp on most gpus with GPUz. It might be called VRM/Vreg/etc something along those lines.

If that doesn't work try Everest.
June 27, 2012 2:44:36 PM


Well i couldn't check the temps for VRM chips with those 2 programs but nevertheless.I want to ask if putting new thermal paste to the GPU will it have any effect or not cause a friend told me it might help.And if this doesnt do i will try to RMa my video card and get a new one. :) 
a b Î Nvidia
June 27, 2012 3:37:01 PM

I'm fairly certain reapplying thermal paste completely voids your warranty, I would just RMA it and hope that fixes it at this point.
July 22, 2012 8:55:26 AM

Sorry i was on vacation for a couple of weeks and couldnt get back to ya.
I couldnt RMA the card cause those morons didnt let me.(i hate them)
But,im thinking of buying a new card ATI or nvidia,amd i dont care just to be a working and around 250 dollars.
My Mobo is:ASRock N68-S3 UCC
Could you give me an example of a card that will fit with the Mobo.
I just want it to run BF3,LoL, and other games mainly BF3.
Thanks :) 

P.S. Could someone explain to me what importance does the Mobo have for gaming and can it be the problem??