Hello, all.
I switched from MS to Debian about a month ago. Have had pretty smooth sailing so far, just a few bumps in the road; mostly Solved by reading forums. The big one that I am facing now is that when I do too much the machine shuts down. When I say do too much , I mean something that in my opinion uses a lot of memory, ie.. transferring gigs of data via wire (it seemed like as the Xfer speed approched 10 mbps) the Computer would, without warning shut off. Using wireless the speed was throttled and did not produce problem. OK that's the first instance.
When I tied compiling a kernel with MTRR and HIMEM support it crashed a couple of times.
I finally got MTRR & HIMEM going and proceeded to finish installing xine . when I did xine-check it said I needed to reinstall xine-lib . Tried apt-get, it did not find it. Here is where I really F-ed up... added unstable to my sources.list and did apt-get upgrade. I knew I shouldn't have done that as soon as I pressed enter. So after rebooting and confirming that I hosed my installation, I tried to figure out how to undo it. After researching I Concluded that a fresh install was less of a hassle. Did that. Now I am back to no MTRR & HIMEM in the kernel. I was looking at some hi-res pictures last night and the damn thing Shutdown again. This time it did not come back up. From a cold start the display was not initialized, no PC beep, nothing, nada, Zip. CD & DVD drives lit up like normal. HDD stayed lit for about a minute.
Removing memory and booting gave the long error beeps. shutdown and reinstall memory and it comes up. Is this a result at no HIMEM Support?
2-512 Corsair sticks in a Soyo Dragon MB, ATI AIW 8500 DV,2.4.27-3-386( i think,not in front of it presently) kernel
Problem has been replicated in various ways, with removing VID card and booting, then reinstalling & booting working a couple of times. If I Shutdown system cleanly then no problems rebooting. My logical hypothesis would be HW problem, but plugging another HDD with w2k eliminates problems.
what gives, Gurus?

I have seen a few computers that just turn off for no reason, and in all of the cases the computer turned itself off because it got too hot. The vast majority of laptops (and a good number of desktops) have thermal shut-off systems that will cut the power if it gets too hot.

The best way to test this is to see if the CPU (or any other component) is hot after the computer turns itself off.

Not sure why this would happen on Linux and not Windows, however.

You hit the nail on the head! I can't believe I did not see the Simple solution. I try too hard to make something out of nothing! Duh. Took the heatsink off of the CPU and could not hold it in my hand for 5 minutes. Separating the fan showed a layer of dust making the fan useless, removed that and no problems to report. I have to apologize, I was too quick to chalk it up to Linux, I am humbled.
Thank you kind sir,
Zach