Yikes! Did Triumph borrow their graphics engine from the Aarklash team?

My GPU fans started screaming after a few minutes of play, even though I had the Vsync and "Wait for GPU" buttons checked. I turned off SSAO as per the suggestion in the official forums, but the temperature still climbs into the 80s from any amount of extended play.

My computer does not overheat when playing Witcher 2, so it's not a case airflow problem.

Running on a Geforce GTX 570 on Windows 7 64 bit.

I'll assume that you've already checked for updated drivers; in the control panel's 3D settings, use Restore to ensure that everything is set to default/application-controlled values. (If you have program-specific settings, get rid of the AoW3 profile too so that it goes back to using the general settings.) I don't know whether it's the case here, but driver-forced settings can cause problems.

Yes, yes I've done all those things. Let me again repeat that other graphically intense games do not currently overheat my system. This problem is unique to AoW 3.

Unfortunately, while reading through their forums I saw they announced they're not going to attempt any more fixes for the problem: Thank god I bought this game at the $15 discounted price! If they're not going to patch their game, I'm not going to buy their DLC. When Eador MotBW runs better than your game that's GOTTA be embarrassing!

Not to quibble, but they are still patching and improving the game. Not addressing overheating issues that plague a very small minority of users is a reasonable cost/benefit decision so that those resources can be applied elsewhere. I'm not saying I like the decision and especially wouldn't if I was a GTX 550 user with heat spikes that paid hard earned money. :)

Yeah, we've all seen Fight Club.

I dunno about this "very small minority" though, a simple google search for "Age of Wonders 3 Overheat" brings up lots of threads on multiple sites.

What's irritating about those threads is even though the OP usually points out the other games that don't overheat, suddenly dozens of the fanboys chime in with, "LOL YOUR CASE AIRFLOW sUx0rZ" and suddenly the whole thread degenerates into arguments over who has the cleanest tower instead of anyone trying to isolate the bug.

Yes, I doubt that we are a "minority" among the players. Rather people that have problems stop playing, like I did. My hardware is update and I own a many gpu&graphic demanding games but any requires that effort to my PC fans.

While it might be comforting to think that all hardware environs are equal and none of them should theoretically ever overheat, that is unfortunately about 180-degrees removed from the facts...;) Especially in laptop country, some hardware combinations are definitely not designed for the rigors of 3d computer gaming, and nVidia has recalled GPUs before for their overheating problems--inabilities to run at their "stock" clocks without overheating under the right conditions, etc.

The fact is that nothing stresses hardware like 3d gaming, and different games stress the hardware differently, depending on how the game addresses the hardware. For instance, years ago when I used to overclock cpus and gpus, I noticed that while certain games ran splendidly during GPU overclocks, other games would crash the system at the same GPU clockspeed. Setting clockspeeds back to default usually always fixes the problem--except in one particular situation, I found: and that is the case wherein a given GPU ships at a clockspeed that is in fact too high for 100% of the software it might be asked to run.

Usually, this happens only during a mistake in the binning & verification process that all chipmakers engage in. Chips which cannot clock as high as others in the same batch & pass verification at the same time are usually either discarded or else *clocked down* until they can pass verification and are marked down in price and sold as a different product . Some of them slip through the cracks and either should have been discarded or else they need to be clocked even lower in order to function correctly in all situations.

nVidia is the only GPU maker that I can recall such "mistakes" happening with, actually. I once briefly owned a nVidia GPU--I believe it was a TNT2--which clocked @ 150MHz stock. The minute I set the clock to 151Mhz (discovered through long trial and error) the GPU would crash when asked to run a 3d game. At 150Mhz it would run all day long--but could not stand even 1MHz of overclock...;) I kid you not. I never saw that again with any GPU...;) But it does happen.

My advice: underclock the GPU to the degree necessary that it takes to run a particular game without overheating--and if that won't work then you've simply got a defective GPU and should replace it under warranty or else buy a new one--in the case of a desktop. Laptops are far more limited in what can be done apart from warranty replacement.