Overclock.net banner

1 - 20 of 25 Posts

·
Registered
Joined
·
161 Posts
Discussion Starter · #1 ·
Ok so here is the deal. I've played, Crysis, Metro 2033, MW2, warhammer 40k, dragon age, etc. All was well until I started playing Splinter Cell (1st one) & pandora tomorrow.

When i would play the first Splinter cell, the game would periodically freeze for a few seconds the screen would go black then pop back up normal. I found out that when this happened all the clocks went to zero for that black out period.

Now I am really worried because the same thing happened to me on Pandora tomorrow except that this time my computer frooze and the GTX 570's fan went to 100%. I could not do anything but hard reset. No respones on the Num Lock LED.

When the computer boots up the card goes back into like 70 to 80% fan speed. Only until a minute or so the video card cools back down. Now if you refer to something wrong.jpg, as you can see when the card goes into 3d mode settings the temp skyrockets for no reason. and the second picture shows metro 2033 running at 100% fan and the temp keeps rising until i closed it out.

Could my card or mother board just crapped out at some point? Or is there something REAALLLYY wrong with playing the old splinter cells. I don't want to RMA this only to break another because of this game.

EDIT:

Ok, I RMA'ed the card and received my new video card today. Here is the tempature profile that it should be. (or was at) Something definitely happened to that card. I wish I knew what happened for my knowledge and for OCN's knowledge. Both attachments with the gpu loaded and high temps are the result of running Metro 2033 on the first stage intro.



 

·
Registered
Joined
·
161 Posts
Discussion Starter · #3 ·
Sorry I did not specify.

I have the EVGA GTX 570 and everything is stock on the video card. Yes about 2 weeks ago the dust was cleaned out of the entire case.

I will update with kombustor tomorrow.
 

·
Registered
Joined
·
2,559 Posts
try forcing the Fan to 100% and see what happens...

i know it will be loud... but still....

also watch the temps
 

·
Registered
Joined
·
308 Posts
Very very strange... I don't care what case one has or what game one is playing, but temps should never get that high at 100% fan speed. Take the card out, clean it again, and make sure nothing is blocking the fan or is inside the shroud that shouldn't be.

After this, reseat it of course and update drivers.
 

·
Banned
Joined
·
82 Posts
either you live on the sun or they forgot to put TIM on your card. just take off the cooler and take a look at the slop application they did
 

·
Registered
Joined
·
60 Posts
Quote:


Originally Posted by pon
View Post

either you live on the sun or they forgot to put TIM on your card. just take off the cooler and take a look at the slop application they did

but it WAS fine... Perhaps the BIOS on the card is on the fritz? I'd try reflashing that.
 

·
Registered
Joined
·
161 Posts
Discussion Starter · #11 ·
This is not something that has degraded with time. It instantly started happening after that last freeze and hard reboot.

I did have an idea about the bios getting screwed up somehow. I will look into reflashing the bios.

as for EVGA OC scanner. The temp literaly shoots up from 50c to 80 c in about 3-5 seconds. I stopped it when it hit 90c. Is there any indication of the motherboard crapping out and somehow causing this?
Quote:
Neopower...550...Only a 550 watt psu?!
Lol I know. From what i can tell the PSU seems to be handling this card fine. I've played many hours of both Crysis and Metro 2033 without issues. I tried cutting cost as much as possible with this build and if i bought new components it would be for a whole new system. The PSU is rated at only 2A less than the required 38A.

EDIT: Ambient temps are anywhere from 23-28c.
 

·
Aren't this fancy!
Joined
·
7,223 Posts
The only time that I know temps shoot up that much is when your benchmarking. So that is actually normal for it to shoot up that much that fast, the program is stressing it to the max instantly.. That's good to test OC's and make sure they are stable at a safe temp.

As for your sudden problem, is it only with those 2 games listed, or now all games are causing it to freeze up?

Actually, this may just be how the card runs, hot. Or like someone else mentioned terrible TIM application. It was only at 60% fanspeed when you hit 89C. Try setting it at highest fan speed, locked, then run something again, benchmark/game and see where the temp caps out. If you don't like the max temp, you need better cooling or check the TIM and see if it was applied alright.
 

·
Registered
Joined
·
161 Posts
Discussion Starter · #13 ·
Quote:
The only time that I know temps shoot up that much is when your benchmarking. So that is actually normal for it to shoot up that much that fast, the program is stressing it to the max instantly.. That's good to test OC's and make sure they are stable at a safe temp.

As for your sudden problem, is it only with those 2 games listed, or now all games are causing it to freeze up?

Actually, this may just be how the card runs, hot. Or like someone else mentioned terrible TIM application. It was only at 60% fanspeed when you hit 89C. Try setting it at highest fan speed, locked, then run something again, benchmark/game and see where the temp caps out. If you don't like the max temp, you need better cooling or check the TIM and see if it was applied alright.
No. The fan was at 100% when it hit 89c. This has NEVER happened. The highest temp I have ever got for benchmarking and gaming was about 72-74. My custom fan profile is set to 90% at 75C and 100% at 80C. With gaming/benchmarking it never hit 90% fan once.

In addition, with the first pic, this shows at NO LOAD 3d voltages (.988V). The temp goes up to 62c with 70% fan. Something clearly happened.

As far as flashing the bios of this evga 570. I cannot seem to find an EVGA GTX 570 70.10.17.00.03 bios. I feel a little iffy on doing this. Will flashing the bios void warranty?
 

·
Registered
Joined
·
161 Posts
Discussion Starter · #14 ·
Something else i noticed. While playing Splinter cell pandora tomorrow. The GPU voltage cycles from .912 to .988. I would say its about 70%low to 30%high of any given time frame.

The GPU load will go anywhere from 5% to 40% load. Even the periods where the GPU load stays about 30 to 40% the GPU voltage is still at the 2d voltage of .912.

Could this have been caused by the GPU rendering at an insufficient voltage?
 

·
Registered
Joined
·
768 Posts
Doesn't fermi have problems with older games? It didn't see sufficient load and would stay in it's idle voltage and clocks? I remember someone joking about running Crysis in the background in order to trick the card out of it's idle state. I also remember something about low GPU usage. On another note, your power supply is fine. It has three 12V rails, 18A from each rail with a maximum output of 42A all together.

The maximum temperature for a GTX 570 is 97C so you are getting pretty close but I've seen several people with fermi cards go that high without issue.
 

·
Registered
Joined
·
161 Posts
Discussion Starter · #17 ·
Quote:


Originally Posted by donnybrasco
View Post

Ok so here is the deal. I've played, Crysis, Metro 2033, MW2, warhammer 40k, dragon age, etc. All was well until I started playing Splinter Cell (1st one) & pandora tomorrow.

When i would play the first Splinter cell, the game would periodically freeze for a few seconds the screen would go black then pop back up normal. I found out that when this happened all the clocks went to zero for that black out period.

Now I am really worried because the same thing happened to me on Pandora tomorrow except that this time my computer frooze and the GTX 570's fan went to 100%. I could not do anything but hard reset. No respones on the Num Lock LED.

When the computer boots up the card goes back into like 70 to 80% fan speed. Only until a minute or so the video card cools back down. Now if you refer to something wrong.jpg, as you can see when the card goes into 3d mode settings the temp skyrockets for no reason. and the second picture shows metro 2033 running at 100% fan and the temp keeps rising until i closed it out.

Could my card or mother board just crapped out at some point? Or is there something REAALLLYY wrong with playing the old splinter cells. I don't want to RMA this only to break another because of this game.

EDIT:

Ok, I RMA'ed the card and received my new video card today. Here is the tempature profile that it should be. (or was at) Something definitely happened to that card. I wish I knew what happened for my knowledge and for OCN's knowledge. Both attachments with the gpu loaded and high temps are the result of running Metro 2033 on the first stage intro.

I hate to say it, but I am having this, what appears to be, the same problem AGAIN. refer to the two pictures i posted. Keep in mind I now have a new motherboard (same model from RMA), and a new graphics card RMA'ed through EVGA.

I was just copying some files across drives when everything stopped responding for a good 3 seconds. The screen then went black and popped back up with a message from windows saying the display driver stopped responding. This seems to be the same problem as with the original card just now it is in the desktop. No temperature problems yet. I want to believe it is the power supply but the card is drawing trivial amounts of current in 2d power saving mode. Could it just be this crap model of motherboard?

I would like to fix the problem asap before something breaks in the video card again. Please, any ideas?

 

·
Registered
Joined
·
129 Posts
Well reading through your post, you haven't eliminated a obscure software/OS issue.

Here's what I would do. Use a old hard drive as your primary and install a fresh OS and few games (saves nuking your current os). Test for a day. If no spare HDD just bite the bullet and reinstall the OS fresh.

Logically beyond that you are down to power supply, maybe test ram.
 

·
Registered
Joined
·
563 Posts
Before install did you run driver sweeper?
Look around for guides on removing graphics drivers here on OCN.
Run CCleaner through registry when your finished, and install latest graphics driver again.
 

·
Registered
Joined
·
2,413 Posts
Quote:


Originally Posted by donnybrasco
View Post

I hate to say it, but I am having this, what appears to be, the same problem AGAIN. refer to the two pictures i posted. Keep in mind I now have a new motherboard (same model from RMA), and a new graphics card RMA'ed through EVGA.

I was just copying some files across drives when everything stopped responding for a good 3 seconds. The screen then went black and popped back up with a message from windows saying the display driver stopped responding. This seems to be the same problem as with the original card just now it is in the desktop. No temperature problems yet. I want to believe it is the power supply but the card is drawing trivial amounts of current in 2d power saving mode. Could it just be this crap model of motherboard?

I would like to fix the problem asap before something breaks in the video card again. Please, any ideas?

I was having the same problem with the Nvidia 260.99 drivers and Dawn of War 2. The game would freeze, the screen would go black for a few seconds and then the screen would come back. Sometimes the screen would go black and stay black and I'd have to reset. I read that it's the driver crashing.

Since you had to hard reboot without shutdown, I would recommend first running sfc /scannow. Then follow that up with chkdsk /r. You can run them from the command prompt.

Google around and see if there is a driver issue with Splinter Cell. Lastly, certain 4xx / 5xx Fermi cards simply do not work (for some people) with certain motherboards. You can check compatibility on this list:
http://forums.nvidia.com/index.php?showtopic=187401
 
1 - 20 of 25 Posts
Top