GPU Overheating

[zalewapl]

With Triple Buffering on, I can stop Grimrock from locking up my entire system, without it on after a couple minutes it'll just nuke the whole thing. That said, if I have Triple Buffering on and run too many intensive programs, it also locks up my system. Now, this is a problem because I have PLENTY of other things that can push my system like this, to the point of moving at a crawl if I run too much, but NEVER has something caused my system to lock up like this.

So it's definitely a fault in the game, not just a hardware problem.

download Furmark from my post above and run the benchmark tests and see if that locks your system; or you start to experience pixel errors or textures errors etc
People need to get out of the mindset that just because a system runs one program correctly it will run another the same way!
Its rubbish!
even if two games are programmed using the same game engine chances are the designers have recoded portions or accessed different ways of creating visual effects which will change how the games play on the same system!
the ONLY thing you can conclude by a system running (for exemplar) battlefield 3 and not crashing is that "in general" its a stable system.
what you dont see is that battlefield 3 might not access memory areas on the video card in the same way as LOG, it might not use the same technology, might not stream pixels the same way...any number of variations that might cause crashing that you might not always see in other applications.
Because LOG can hit your GPU hard it might be exposing underlying issues that other system software just hasnt exposed!
benchmarking with Furmark will burn your GPU good and i promise you if its going to lock up under stress than the benchmarks in furmark will cause it to do so!
for extra stress testing try unigine heaven:
http://unigine.com/products/heaven/
Heaven is great for stress testing overclocking as it will very quickly display white pixels of texture corruption if youve overclocked your card too much.....hitting higher settings and enabling tessalation (if you run Windows 7) will also hammer your card for a decent stress test.
remember LOG isnt a standard 2d dungeon crawler..its actually a fully fledged 3d enviroment (right clicking and holding your mouse and looking around will prove that) so it will be using alot of gpu power even though its not flashing fancy effects all over the place!

again...stress test with Furmark or Heaven and post your temp results....

That is something I can agree with. I have to add something from myself though:
Furmark failed me once. I was overclocking my graphics card and tests done with Furmark went all OK: the temperatures were in acceptable boundaries, there were no artifacts and the system was stable. However, everytime I launched "Rainbow Six: Vegas 2" and went to one specific spot on one specific map the entire system immediatelly blue-screened. I could reproduce this every time and it only stopped happening once I lowered the overclocking parameters. So I have to say that Furmark is not 100% reliable.

On Grimrock:
The game also raised the temperatures on my GPU to unusally high values - up to the point when the fans became noisy. But my system didn't crash/blue-screen/freeze/lock-up/do whatever when that happened and it never does. I could still play the game normally and I only noticed the noise from the fans when I put away my headphones. The temperature issue was fixed when I enabled vertical sync in the game menus.

When you buy PC equipment it is not a matter of purchasing the highest possible parameters for the lowest possible amount of cash. It's an investment. It pays off not to skimp.

Long story short:
When your entire system locks up, shutdowns or goes to a blue screen when you're playing a game it's never the game's fault. The culprit is always one, or a combination, of following: faulty hardware, insufficient cooling, faulty drivers, non-updated BIOS or a non-updated operating system.

[Nubcakes]

On Grimrock:
The game also raised the temperatures on my GPU to unusally high values - up to the point when the fans became noisy. But my system didn't crash/blue-screen/freeze/lock-up/do whatever when that happened and it never does. I could still play the game normally and I only noticed the noise from the fans when I put away my headphones. The temperature issue was fixed when I enabled vertical sync in the game menus.

When you buy PC equipment it is not a matter of purchasing the highest possible parameters for the lowest possible amount of cash. It's an investment. It pays off not to skimp.

Long story short:
When your entire system locks up, shutdowns or goes to a blue screen when you're playing a game it's never the game's fault. The culprit is always one, or a combination, of following: faulty hardware, insufficient cooling, faulty drivers, non-updated BIOS or a non-updated operating system.

You obviously never ran Furmark. Furmark makes video cards run FAR hotter and draw more power than they ever could under normal circumstances. Most computers will overheat and burn something out if they run furmark for more a few minutes.

[zalewapl]

You obviously never ran Furmark.

How can you say I didn't when I explicitly said in my post that I did? I said that "temperatures were in acceptable boundaries" but that might have been a bit brief. I meant that the temperatures were well above any game would ever raise them, but they were still below the specified by the vendor maximum threshold and I experienced no problems at all.

Most computers will overheat and burn something out if they run furmark for more a few minutes.

Most likely they will shutdown before anything gets permanently damaged because that's how hardware works these days - it's designed to protect itself. Unless you buy the cheapsest (spelling intentional) stuff.

[Nubcakes]

You did not say how long you ran furmark for. Unless you specifically state which test you run, its assumed you did the 1 minute benchmark. Try running the BURN-IN test using the "Xtreme burn-in" setting at 1920X1080 @ 8X AA for 24 hours and let me know how that turns out!

I work at a computer service shop and one of my co-workers decided to see what would happen if he let a computer running a GTX560 run furmark for as long as possible. The computer was not in a case and the brand of the GTX560 was eVGA which is not really on the low-end of the quality spectrum. The ambient temp was around 65F. After about 4 minutes the card temperature hit 107C and started showing artifacts, The temperature didn't exceed 113C, but after 20ish minutes the card started smoking and the computer crashed. Everything seemed to be fine except the videocard which wouldn't allow any computer it was in to POST. For who dont know; POST stands for Power On Self Test. It's the screen you get when you first turn on a PC, assuming you don't get a splash screen for some company(EXA: Compaq, Intel)

This experiment was redone with a XFX Radeon HD6970, the end result was the same. The temperatures and times were slightly different but the point still stands;

You didn't specify how long you ran furmark for. I'd wager any computer could run the 1 minute bench mark no problem. Running a stock air cooled videocard for hours on end? Not so much.

Edit; for the record I use 2 GTX560 TI in SLI and the normal gaming temperature I see is 63C on KillingFloor and The Witcher 2 (Games I am playing right now). Legend of Grimlock runs them at 59C. So I actually don't see the problem most people are complaining about in this thread.

[labecula]

No, no overheating here. Temperature ist as normal as in other Games like this. Not at peak like BF3 or so.

2600K
GTX570
Win7 x64