Certain graphical events make my computer think it needs to shut down

[Suiram]

A while back I had an issue where certain animations like escape, warp, firaga, etc would make my computer power down. I have a Dell XPS M170 laptop, and when I ran the Diagnostics thing at the time there was an issue with one of my fans. I replaced the fan, which stopped the shutdowns, and then got newer drivers for my video card (nvidia GeForce Go 7800 GTX), which caused there to be less stuttering for those kinds of events.

However, after some time I noticed that I would get similar shutdowns whenever there was earth weather. So I filtered weather effects, and the problem stopped. I had even forgotten about it until last week, when I was in Sauro on a mule that I hadn't filtered them for and got shutdown. At the time I ran the Diagnostics to see if there was a new fan issue, but both fans are working properly. Then today I was doing WotG quests, which have dust/earthy animations prevalent in many of the cutscenes, which I can't turn off. If anyone has done these quests, you'll know there's one in particular for Windurst that's basically 10 minutes of dust, explosions, and fire.

Basically my question is this. I don't think my PC is actually overheating, there's no significant blockage of air (and I used a compressed air can on it anyway). But for some reason certain graphical events make my PC think it's overheating and it shuts down. I have the newest drivers for my video card (from January 08) and my BIOS (some time in '05). Is there anything I can do just to simply overwrite the automatic shutdown thing?

[Kohan]

Although you believe your laptop is not overheating, all signs point to that. When the load is especially high, the components are naturally generating more heat, and it's only when things are at their peak that your computer shuts off. Considering that you have a high-end Dell laptop, you wouldn't expect this to happen, but that's a compact space you're dealing with and an on-board chipset -- it's very possible.

Here's an application that can help you determine internal heat: Core Temp

However, if we were to get into other potential culprits, here's one: defective power supply. Unstable power could possibly be causing shutdowns when your machine is demanding more of it but the right amount of energy isn't being delivered. Naturally, this would mean you'd need to replace the board in your laptop, if it were the problem.

I'd suggest changing drivers, but you've already updated that, and you updated your chipset drivers too from the sound of it. Really, though, I don't think it's related to those.

You can also try running this with a "Blend" test: Download Stress Prime 2004 BETA

If your computer turns off after that, then it's not related to your graphics chipset, but your CPU and/or RAM.

[Suiram]

It seriously only happens when there's "dust" rendered in FFXI. After I made this post we went and did a different quest, and everything went fine. Then in the cutscene after we won the orcs and NPCs started kicking up dust (it really was a gratutious amount of dust, designed--I'm convinced--for the sole purpose of causing my computer to crash). But anyway, once I saw the dust I remarked to my friend in the cutscene with me that I was going to crash, and low and behold it happened again about two minutes in.

Now, of course, I suppose it's possible that the dust effect is actually overheating my computer, which is in turn responding normally, but I've run my PC before for a long time in high-lag areas in the summer, and I know how it feels (temperature-wise) when it gets hot, and it was luke-warm when I felt that bottom panel. Anyway, I'll check out those links you posted but I thought I should update with my recent experience. Also, I didn't actually update my chipset, because I checked and I was running the latest version.

[Kohan]

Touch, unfortunately, does not tell you enough. A mere percentage or two may be all it takes to cause a shutdown, and you will not be able to tell the difference with your hand. I understand your frustration, and that you've felt the machine running hotter before, but that still doesn't completely rule out this as a possibility. All it takes is a slight jump, and that's it, shutdown.

This can also become more problematic over time, which explains why you have been able to alleviate/avoid it in the past, but it has returned. As components age, problems increase, whether it's due to things becoming unseated, spread further apart, et cetera.

However, the other problem aforementioned -- problems in the power supply -- is very possible too. Unfortunately, without the proper tools to open it up and safely test whether it's putting out what you need, you can't easily verify this.

Also, if you determine that it is not your GPU chipset but rather another component overheating, you may be able to enter your BIOS and change some settings that will prevent your machine from shutting down. This could complicate things in the future and/or damage your components further if things run out of control, though.