Sign in with
Sign up | Sign in
Your question
Closed

Nvidia GTX 570 crashes PC while playing games

Last response: in Graphics & Displays
Share
June 6, 2012 2:33:31 AM

Whenever I try to start up DOTA, SC2, D3 etc my PC automatically black screens and crashes/freezes 10-15 seconds after having a game up. Usually I can't even get past the log-in screen.

Here are my specs:

GPU: NVIDIA GeForce GTX 570
Processor: AMD Phenom II X4 Processor (4 CPUs), ~3.4GHz
Memory: 3584MB RAM
System Model: TA790GX A3+
PSU: Thermaltake TR2 800W
Windows 7 Ultimate 32-bit (6.1, build 7601)

I'm running the latest GTX 570 driver (version 301.42).

I have no idea what to do. I previously had an ATI graphics card on my pc before I switched over to Nvidia. When uninstalling my ATI software I booted into safe mode and ran driversweep and removed off ALL ATI affiliated things just to make sure my Nvidia card would run smoothly... apparently that wasn't the case.

Edit: Here's a screenshot of my temps and voltages.. Does anything seem off/irregular?

a c 99 U Graphics card
a b Î Nvidia
June 6, 2012 3:37:46 AM

How long have you had these issues?

Try the Beta drivers. If not roll back to 29x.xx drivers.
Score
0
June 6, 2012 3:43:00 AM

I've had these issues ever since I got the card. Was previously on 29x.xx's.. didn't work. Which beta's would you suggest?
Score
0
Related resources
June 6, 2012 3:54:32 AM

Quote:
Might be your PSU going bad.

Do you have another PSU to test with?


I don't think it's the PSU as I was running an ATI Radeon 5750 before this new card without any issues at all. I can switch it out with a spare 550watt PSU but I doubt it would be enough to run it.
Score
0
June 6, 2012 4:55:05 AM

Quote:
The 5750 uses far less power than a 570.
The 550W will be fine as long as it is a decent brand.

If not, you can power the GPU seperatly with only the 550w and use your 800 for the rest of your system, this should help narrow down the issue.


Would you be able to suggest any other possibilities? I highly doubt it's a PSU issue to be honest. Which drivers would you suggest I roll back to if any?
Score
0
a c 99 U Graphics card
a b Î Nvidia
June 6, 2012 5:03:51 AM

You can RMA it, if you still have warranty. Since you mentioned, "ever since I got the card."
Score
0
June 6, 2012 11:12:55 AM

Sorry to hear your having a problem Cepsbrahh

Quick question, what is the brand of 570 your using (asus, msi, evga etc)?

I too had that problem with a GTX 570 that I bought a while back, my system would literally shut off minutes into starting up a game.

Turns out it was a faulty card, allowing the temperatures to shoot up into the high 90's and even to 100c. Obviously the overheating caused the PC to shut down everytime to avoid more damage.

Id suggest that you download GPU management software (preferably from the company that make your card) such as EVGA precision or Asus Smart Doctor and monitor your temperature both on the desktop and when you start up a game.

If yours too turns out to be a faulty heating issue, return it and get it replaced.

Note; it may be worth trying to quickly take a print screen of the temps before you shutdown, it will help in the email youll have to send to your supplier.
Score
0
June 6, 2012 1:45:21 PM

What temp monitoring software would you suggest? I'll try swapping out the PSU once I get home today.
Score
0
June 6, 2012 1:57:46 PM

As I said before, the manufacturer of you card will usually have their own software available to you.

I have an EVGA 580, so I use EVGA Precision to monitor by Temps, fan speed etc.

But there are general programs like the ones malmental mentioned above.
Score
0
June 6, 2012 3:38:32 PM

Could it be possible that the card's not fully seated properly? It kind of awkwardly went in the slot at first; almost had to force the thing in.
Score
0
June 6, 2012 5:22:32 PM

Usually if the card isnt seated properly, it wont work at all rather than not work when under load - but it would be a good idea to reinstall it if there are any doubts.

Seeing as its fine under normal conditions but under load (such as playing a game) it your system shuts off really does scream Temperature issue to me.

Please update this post once you tried looks at your GPU temps, Im interested to know if your another 570 owner who had my problem :) 

Note; Up to 45°C is ok for idle, up to 80°C is ok under load. 85-100°C will shut you down (from what I remember anyway)
Score
0
June 6, 2012 5:41:16 PM

Gallarian said:
Usually if the card isnt seated properly, it wont work at all rather than not work when under load - but it would be a good idea to reinstall it if there are any doubts.

Seeing as its fine under normal conditions but under load (such as playing a game) it your system shuts off really does scream Temperature issue to me.

Please update this post once you tried looks at your GPU temps, Im interested to know if your another 570 owner who had my problem :) 

Note; Up to 45°C is ok for idle, up to 80°C is ok under load. 85-100°C will shut you down (from what I remember anyway)


Do you really think it could raise that high in temperature that fast though? As I said, my PC would shut down from 10-15 seconds of activity on the log in screen ALONE. Don't even think the log in screen requires significant graphical processing at all. I'll give it a shot and monitor the temps though.

Do you think all driver issues are ruled out since the games actually boot up?
Score
0
June 6, 2012 6:50:08 PM

I had someone from a separate forum suggest me this fix:

Quote:
easy fix same thing has happened to mannny ppl since newer drivers...install msi afterburner go into settings and tick allow voltage change then on the main screen type 1063 in voltage hit apply, problem gone. the tiny lil bump well fix your issue I bet and you ned to do this all the time or set msi to do it when pc starts.


Does this seem like a possibility? Looks as if the card's pre-programmed to request an astronomical amount of voltage causing the PSU to crash.
Score
0
June 6, 2012 6:52:22 PM

Quote:
MSI afterburner works a treat if you need to change fan speeds and dial back voltages.


Do you think that the card's simply demanding too high of a voltage causing the PSU to crash?
Score
0
June 6, 2012 8:57:21 PM

cepsbrahh said:
Do you really think it could raise that high in temperature that fast though? As I said, my PC would shut down from 10-15 seconds of activity on the log in screen ALONE. Don't even think the log in screen requires significant graphical processing at all. I'll give it a shot and monitor the temps though.

Do you think all driver issues are ruled out since the games actually boot up?



Yeah trust me, when I had this issue, I would literally get into a game and few seconds later - shut down. Its surprising how fast a GPU will heat up if it doesn't have any cooling. Eventually it would just start doing it on the desktop.
Score
0
June 7, 2012 4:39:39 AM

Just recorded the temps and voltages of all my PC's components. I assume the values under the "max" column are at the time of the crash.

Score
0
a c 99 U Graphics card
a b Î Nvidia
June 7, 2012 5:44:53 AM

Yea, I find it interesting as well. But, I'd still suggest an RMA.
Score
0
June 7, 2012 12:38:43 PM

Quote:
interesting....


Care to elaborate why? Do the numbers seem normal? Would they suggest it's a defect with the card itself?
Score
0
June 7, 2012 12:54:37 PM

Hmm thats strange... in a way Id have liked your issue to be temps so you could quickly RMA the card and be sorted, but this makes it a little more difficult to diagnose.


If you havent already, I would suggest contacting the supplier of your card with the issue and see if theyre ok with taking it back, or if they offer any suggestiosn themselves.
Score
0
June 7, 2012 4:24:15 PM

Quote:
so we have to go back to the power supply and/or the graphics card itself..


Although I highly doubt this is the issue I'll give it a shot. Some guy suggested I re-seat the card, too.
Score
0
June 13, 2012 9:24:52 PM

Quote:
as of now which drivers are you running.
and if you install new of different drivers please select the 'clean install' option under 'custom install'..

re-seating the card couldn't hurt..


The same driver listed in the OP.

I tried re-seating my card off of the "Slave" slot and onto the "Master" slot. Same issues.
Score
0
June 14, 2012 8:51:57 PM

I installed a fresh x64 bit windows 7 last night and installed the v301 drivers again. nothing has crashed yet. I'll install Starcraft 2 as soon as I get home and I'll try it out. I'm hoping it was a registry/driver issue with my last build, because I DID have an ATI card previously installed on it.
Score
0
June 15, 2012 1:05:53 PM

Quote:
just might work...


Installed Starcraft last night and it ran for about 15 seconds before my PC crashed... the rebuild in 64 bit did nothing.

I've narrowed it down to my motherboard or the card itself. Do you know what the most secure driver version would be? I think it's best not to be testing with the latest version.
Score
0
June 15, 2012 1:14:42 PM

Quote:
I think you need another card (GPU) to try out.
all other factors have been checked it seems.

I'd say GPU before motherboard...


Do you have any idea what the most stable driver version would be? I think it would be best for me to not do all this testing with the latest version as Nvidia drivers have always been known for glitches/defectiveness with latest releases.
Score
0
June 15, 2012 1:39:03 PM

Is your PC crashing when it locks up going into a game? Are you then logged back into windows and pulling up your temps and voltage? because all those values are NOT under load. Those are normal boot up / idle temps.

Download Furmark and have the GPU monitoring software open while you have furmark running in window'ed mode. This way you can see on the fly what your temps are. Rather then exiting or rebooting your pc and opening another application to see what it currently is.

No video card goes up only 2C under load. But according to your image, your 570 idles at 37C and on load goes to 39C... no way.

Score
0
June 15, 2012 1:44:53 PM

I see you have yet to swap out your PSU.. that would of been number 1 to test. Number 2 would be, toss back in your old ATI or another video card.

You probably would of been solved by now. Doing nothing more then ^ those 2 steps.
Score
0
June 15, 2012 1:48:41 PM

Triton05 said:
Is your PC crashing when it locks up going into a game? Are you then logged back into windows and pulling up your temps and voltage? because all those values are NOT under load. Those are normal boot up / idle temps.

Download Furmark and have the GPU monitoring software open while you have furmark running in window'ed mode. This way you can see on the fly what your temps are. Rather then exiting or rebooting your pc and opening another application to see what it currently is.


When my PC crashes the screen goes black and I'm forced to restart. I grabbed those temps after restarting the PC and firing up the temp monitoring program.

I ran furmark last night and my PC crashed in less than 10 seconds.

Now would this indicate a driver issue over a hardware issue with the card? How could you differentiate? There's no way it's a temp issue; it couldn't possibly overheat in ~7 seconds after running furmark.
Score
0
June 15, 2012 1:52:27 PM

Right, then those temps are absolutely worthless to be posting. Since they are a min or two after the intial crash happens.

If you havent tried to swap out the PSU do that. <<<<<<<<<<<

If you still have the issue with the new PSU, then its the card. Your wasting so much time with all these installs and OS rebuilds, you could plug in another PSU, and just hook up the 6pin_PCI cords to JUST The gpu in less then 5min and have your answer.





Score
0
June 15, 2012 2:01:37 PM

Triton05 said:
Right, then those temps are absolutely worthless to be posting. Since they are a min or two after the intial crash happens.

If you havent tried to swap out the PSU do that. <<<<<<<<<<<

If you still have the issue with the new PSU, then its the card. Your wasting so much time with all these installs and OS rebuilds, you could plug in another PSU, and just hook up the 6pin_PCI cords to JUST The gpu in less then 5min and have your answer.


From what I understood the program records the record high and low. I think the card gets to the max temp listed right before my PC crashes, so that value's probably correct. It seems realistic it would only go up a couple degrees after simply loading up the log in screen of a game, right?

I'll try the PSU swap as soon as I get home. You're probably right about the simplicity of this problem but you never know with Nvidia drivers.... they're always faulty and causing problems.
Score
0
June 15, 2012 2:05:44 PM

He suggested to use my 550W solely for the GPU and the other one for the rest of the system. Haven't tried that yet.
Score
0
June 15, 2012 2:07:03 PM

/facepalm

When your PC reboots, it loses all that cached info.. temps and such. Unless you literally have it spitting out to a log file. Which you do not. What you show is an application that shows CURRENT high / low from the moment you execute/open it.

Nvidia drivers are fine. Sometimes we run into issues. But for the MILLIONS of people that have Nvidia cards. If we all ran into PC crashes cause of drivers. The forums would be filled with HUNDREDS of posts just like this.

You have a hardware issue. Its either the PSU, or the Video card. My money is on it being a faulty card.



Score
0
June 15, 2012 2:14:00 PM

Triton05 said:
/facepalm

When your PC reboots, it loses all that cached info.. temps and such. Unless you literally have it spitting out to a log file. Which you do not. What you show is an application that shows CURRENT high / low from the moment you execute/open it.

Nvidia drivers are fine. Sometimes we run into issues. But for the MILLIONS of people that have Nvidia cards. If we all ran into PC crashes cause of drivers. The forums would be filled with HUNDREDS of posts just like this.

You have a hardware issue. Its either the PSU, or the Video card. My money is on it being a faulty card.


The temps listed under the "Value" column are current temps, Max is the absolute logged maximum temp regardless if your computer crashes or not. It's all logged automatically. It's not as if you have to save it yourself.
Score
0
June 15, 2012 2:27:22 PM

cepsbrahh, you came here for help and advice and you haven't tried the best advice yet. Test your PSU. Don't bounce all over the place with what you think it must be or must not be. Eliminate the possible and then whatever is left is the probable.
Score
0
June 15, 2012 2:52:48 PM

cepsbrahh said:
The temps listed under the "Value" column are current temps, Max is the absolute logged maximum temp regardless if your computer crashes or not. It's all logged automatically. It's not as if you have to save it yourself.


Unfortuantely, you're wrong.

CPUID Hardware monitor resets its values any time you shut down or reboot. So if you have it open, and your PC crashes.. and you power back up > Then log back INTO windows, and then open up CPUID Hardware monitor AGAIN. Its starting a NEW log of current temps and volatages.

Its not going to show you anything prior to your last reboot.

You've got alot of solid advice and many of the members here have been advising the SAME tips. Yet you continue to go off on your own.

Swap out the PSU, then post the results. We can then take it from there.


Score
0
June 15, 2012 6:15:28 PM

Just letting you all know.. I HAVE swapped out the PSU... but I didn't supplement it with the other PSU as well. Had the same problems by the way. I'll pool both PSU's together and dedicate the power of my 550W one solely for my GPU. We'll see if that changes things. If it does.. then I'll probably need a better quality 800W PSU... either that or it's faulty which I highly doubt since my last build/GPU ran perfectly fine without a single crash. That was what led me to believe it was a driver issue in the first place.

I appreciate the clarification of how that temp monitor program works, I apologize for being stupid.
Score
0
June 15, 2012 6:22:24 PM

Youre not stupid! I just dont want mis-information floating about ;) 

I had a GTX480 running on a 600w without any issues. If you tried the 570 on both your 550w and the 800w and they are both doing the same thing ( crashing ) then its the card.

Just RMA it. What you are describing is what happend to me back with a GTX275.

Good luck!
Score
0
June 22, 2012 5:46:27 PM

I found out what the issue is... it's a blown capacitor. One of them blew for some reason, it's sitting on the MOBO slightly bent and the surface of it's black. All symptoms of a busted cap align with the problems I've been having.

Now regarding blown capacitors... how does one go about replacing them?

Are there specific caps for specific motherboards? Should I try fixing it myself or should I take it into a repair shop? I have a solder at home, so lack of tools isn't an issue.
Score
0
August 18, 2012 6:09:09 AM

So this may be an old thread but i recently had the same exact problem, any time under load the entire PC crashed. After reading this forum i thought hey maybe I've got a bad capacitor. So i open up the computer and did two things.

1. My case had two audio options, only one of which was supported by my motherboard, this left a hanging wire which had found it way to resting against the backside of the card. I moved this wire.

2. I also reconnected the two power connections on the card, making sure I'm pushed them in nice and hard, also putting some force on the card itself. However i didnt entirely reseat the card.

Ever since I've been golden, its one of the stranger things I've had a computer do to me. Oddly enough my 570 had worked flawlessly for months until yesterday and it's been driving me insane ever since. So for anyone encountering a similar problem, I hope this helps. Btw so its well known, I've watched my temps on EVGA precision up until the second I crashed. It happened so fast i never even hit 50. RMA is not always the answer.
Score
0
a c 271 U Graphics card
a c 171 Î Nvidia
August 18, 2012 7:39:25 AM

This topic has been closed by Mousemonkey
Score
0
!