Games constantly crashing

[Paliente]

I bought a new system on Friday and got it home all excited and ready to use. Installed a couple of my games and they hard lock the system after about 1-2 minutes of play (cannot alt+ctrl+del, alt+tab or anything but reset the machine). This happens in COD4, Crysis, Bioshock and NFS:Pro Street.

I have installed the latest DirectX:June update, and tried it with 175.70, 175.16 and 177.41 drivers. Turned off Firewall, AV before playing the game. The system I have is an E8400, 4GB Corsair 6400 memory, Galaxy 8800GT (My GPU-Z pic is below), Gigabyte EP45-DSL3 M/Board. I'm also running Home Premium x64 edition that has already come with SP1 installed. PC is not overheating as it happily stays around 35-45 before it crashes. Nothing is OCed and everything is at its default settings.

Any other assistance would be great. I knew I avoided Nvidia for 6 years for a reason :\

http://img.photobucket.com/albums/v652/paliente/GPU-Z.jpg

 

[ChrisJust66]

When you say bought, you mean you actually purchased the computer pre-built, right? And is it always a few minutes after you start the games or does it seem random?

 

[Paliente]

When you say bought, you mean you actually purchased the computer pre-built, right? And is it always a few minutes after you start the games or does it seem random?

Sort of, I chose all the parts I wanted in the PC and the place where I bought it from built it. I'm not all that confident in doing it myself (been out of hardware for awhile now).

The game always crashes after a few minutes of play (essentially the same time passes in the game before it hangs). I could repeat it in Crysis as it always crashed in roughly the same spot each time after running to it. It also hung in Crysis not far into the intro if I didn't skip it (Where the lady archeologist is speaking).

There's got to be some logical explanation for why it would run then crash. I ran the graphics test 3dMark Advantage (for the PC with my specs, the first few tests only ran at 5-10 FPS whereas I would have expected better performance). Could this point to possible compatibility issues with driver and system/card?

Anyone know of some very stable drivers that I could use?

(sorry about delayed reply, different time zones and the fact I work random hours of the day don't help )

 

[BewareOfButtlice]

To me it almost sounds like your computer is overheating or you may not have a good Power Supply installed. What brand and how many watts is your PSU? Did the people who built your computer overclock your processor? Download and install Everest Ultimate Edition - Free software downloads and reviews - CNET Download.com and run a stress test on your computer for about 30 minutes to an hour and let us know the temperatures.

 

[Paliente]

To me it almost sounds like your computer is overheating or you may not have a good Power Supply installed. What brand and how many watts is your PSU? Did the people who built your computer overclock your processor? Download and install Everest Ultimate Edition - Free software downloads and reviews - CNET Download.com and run a stress test on your computer for about 30 minutes to an hour and let us know the temperatures.

Thanks for getting back to me and I'm going to try Everest now but in the meantime the PSU is: EarthWatts 500 Watt power supply (80 PLUS® certified) if that helps. Will post again once Everest is finished. The one thing I don't understand is that there are 2 PCIE plugs going directly from the PSU into the Graphics card. http://www.antec.com/connectors.php?ProdID=15137&SetZone=en has the picture of the PCIE connectors.

For extra information, the graphics card supports PCIE2.0 and the motherboard (EP45-DSL3) is PCIE2.0 too. Would having PCIE1 power cables going into PCIE2.0 graphics make a difference? i.e. voltage isn't enough or too much?

Here's the results of the Everest test:
Not much happened for about 15 minutes and the system would have normally crashed after a few minutes. The reason why the tick boxes were empty then was due to me testing out what each does. Will try searching for a graphics stress tester now though 3d Advantage would do a good job of testing it.

 

[roy69]

Some graphics cards do use two power connectors. If it was installed incorrectly you would know about it.

Sounds like it is graphics or memory related.
Check the drivers that came with the computer.
Some times reverting back to the origanal drivers by the manufacterer will help.

 

[Paliente]

I believe that it could be a heat issue afterall. I've just run a program called Video Card Stability Tester with Everest monitoring the GPU and GPU memory heat and it hard locks the system when the GPU hits 64 degrees (Celsius or 147F) and GPU memory hits 55 Degrees (Celsius or 131F). Does these sound like too hot temps for a graphics card?

My friend's PC had similar cooling issues with his old P4 3GHZ but we later found out that HP had installed a TV card directly in front of the Graphic's heatsink/fan setup which meant that the heat had nowhere to go, the fan promptly melted and the PC rebooted.

I really appreciate everyones help on this matter so far. If this is the case then it looks like I will need to get new cooler installed on the 8800GT. It's pretty pathetic that a brand new semi-top of the range card can't cool itself properly (stupid 5c fans).


It seems to spike quite quickly and this was ~1.5 minutes and the program didn't even get a chance to complete its 3 minute benchmark.

 

[roy69]

The galaxy version of the card does get hot so it is not that problem. A quick google are reporting temps up to 75c

 

[Paliente]

Thanks for doing that Roy.

Just finished a full memtest and it found nothing wrong. Going to take it back to the place that assembled it and make it their problem.

 

[ChrisJust66]

I have 2 EVGA 8800gt's and they IDLE around 89C. Yeah I know, sounds high but whatever. They drop abit when I open the door, my case is a cosmos?, but not that much. If that's your temps, then I doubt it's overheating. I would go with the driver or memory problem. There's a few different things you can try to narrow it down.

1. Take the side panel off for better airflow, allowing you to check for overheating.

2.Take all your ram out and start with one stick, then keep adding until it locks up. If it does it on the first one, try a different one. If it does the same thing on the next one, then it's probably not the ram. You could try another one to be safe.

Make sure you have all the latest drivers to. And do me a favor, that is if you haven't takin it back yet, and let the everest stress test run on the cpu only for about 8 to 10 hours. When you go to bed. I'm not positive on this but it does sound like the problem I had a few months back. But your specs are different than mine. That's why I'm wondering.

 

[Paliente]

I have 2 EVGA 8800gt's and they IDLE around 89C. Yeah I know, sounds high but whatever. They drop abit when I open the door, my case is a cosmos?, but not that much. If that's your temps, then I doubt it's overheating. I would go with the driver or memory problem. There's a few different things you can try to narrow it down.

1. Take the side panel off for better airflow, allowing you to check for overheating.

2.Take all your ram out and start with one stick, then keep adding until it locks up. If it does it on the first one, try a different one. If it does the same thing on the next one, then it's probably not the ram. You could try another one to be safe.

Make sure you have all the latest drivers to. And do me a favor, that is if you haven't takin it back yet, and let the everest stress test run on the cpu only for about 8 to 10 hours. When you go to bed. I'm not positive on this but it does sound like the problem I had a few months back. But your specs are different than mine. That's why I'm wondering.

Already took it back before I read this, I shouldn't have to really stress about these things too much seeming it's so new and all. They told me 2 hours but it's been 6.5 hours since I took it in so they're probably having as much fun as what I had with it. Hopefully it wont be a month before I get it back now. I don't mind problem solving these things but when you fork out $1500 you'd expect it to work out of the box. Thanks for the suggestions anyway Chris.

 

[ChrisJust66]

No problem. I figured it might've been to late but i thought I'd give it a shot. Let us know what it was when you find out.

 

[Paliente]

It's been roughly 3 days and I've not received word of my PC being fixed. They probably don't have a clue either.

 

[ChrisJust66]

That sucks. I'd call them up if you already havne't. If they haven't figured it out by now, their probably going to set you up with a new one. But I'd still like to know what it was if they figure it out.

 

[stoneys-nutz]

just a thought, what video driver are you using ? i know there are issues with the latest driver from nvidia, the 7517, it seems to be causing lock up's among other things use the 7.15.11.7516 and see what happens. tell the repair people

 

[Paliente]

just a thought, what video driver are you using ? i know there are issues with the latest driver from nvidia, the 7517, it seems to be causing lock up's among other things use the 7.15.11.7516 and see what happens. tell the repair people

Hey Stoneys, thanks for the suggestion but when I had the machine I tried 3 different drivers including the one you mentioned (175.70, 175.16 and 177.41).

I haven't really stopped thinking about it since I took it in. My logical brainwaves suggested that once the 8800 gets to around 64 Degrees its fan ramps up or something happens and the power needed to power everything including the new fan speed puts too much pressure on the PSU (500W which should be enough unless the PSU has some sort of issues dealing with large drains). This results in more power being allocated to the 8800, therefore, starves the motherboard (choking the CPU, memory), hard drives or it just can't allocate enough power to supply everything and it all just decides to die. This power starvation results in none of the usual problems associated with overheating which I thought of a few days back. I should have realised this originally as I've seen what happens when graphics cards overheat.

This is all just speculation and stuff I thought about in the endless boring periods at work.

 

[stoneys-nutz]

i think you may be on the right track , i would try another psu, 66ow at the very least. 500w is just about the min required for sli, go here and you can see which psu is required for your system, hope this helps. go to tools and the free version
is on the left.

Link.

eXtreme Outer Vision - eXtreme tools for computer enthusiasts

 

[Paliente]

i think you may be on the right track , i would try another psu, 66ow at the very least. 500w is just about the min required for sli, go here and you can see which psu is required for your system, hope this helps. go to tools and the free version
is on the left.

Link.

eXtreme Outer Vision - eXtreme tools for computer enthusiasts

What a convenient little site. Judging by the result it gave me, 500W should be enough.

Just heard back from them then and they said they replaced the Memory and the video card and it started working, though they did not give me any specifics on what it was. It only took them a week to do this......*sigh*

Though

their probably going to set you up with a new one.

is most likely what they did and the reason it took so long was so they could set it up like how it was before.

Thanks to everyone for their help/suggestions on this, it's great to know there's always people willing to help on the net. This is one site I'll definitely visit often.