I made some hardware upgrades today: installed a new video card, a new CPU cooler, and a DVD burner. Everything seemed fine, but every 10-15 minutes the computer just freezes. Freeze is actually the wrong word since I can still move the mouse's cursor, but it has no effect. Keyboard input is useless and I just have to reboot. On the 3rd reboot fsck checked the root partition and made some changes.

I have checked system/CPU temperature to see if maybe the cooler wasn't working right and checked it's mounts. Everything there is great.

The system hangs whether I'm using the burner or not. (It's an ASUS)

The hangs seem to happen when scrolling a window or changing desktops, but I'm constantly doing something to that effect so it's hard to say if that's the cause.

The card is GALAXY Geforce 240GT. I've never used a Galaxy card so I'm leaning towards that, and have gone back to the old Geforce 9400 in the meantime, but while I'm waiting to see what happens, I was wondering if maybe the filesystem getting modded in the fsck check might point to a filesystem error causing the hangs, or if it seemed more likely that the filesystem was corrupted by the hangs.

Any thoughts?

I'd guess it's the video card. Some manufacturers like to factory over clock their cards, which cause overheating. Or could be bad RAM chips, bad heat sink mounting, cheap fan - on the video card.

The Nvidia settings program will monitor the temps for you. My GT220 after some use (Warzone, SpringRTS) hits 42-44C, with KDE's effects on, it idles at 34-36C.

If the same issue(s) happen with the 9400 back in, we know it wasn't the new video card. I'd also check the PSU (new CPU, more powerful video card = more current draw), run memtest, along with the standard smart tools, and manufacturers hard drive test utility (which I'm sure you already did once fsck found corruption )

It definitely seems to be the card since I watched 30 minutes of some random DVD on the new burner with the old card. .

As far as disk utilities go, I know the disk has issues (bad logical blocks to be exact), but I'm letting it live out the rest of it's life in this testbed. The PSU is a new 500W coolermaster which I trust and is well over what the card said it needed (and well under what the old 9400gt says it needs oddly enough). I actually upsold myself on this card anyway since I was really going in for a EVGA GT220 that's on sale. I imagine I can get a new disk and the card I was after to begin with for just a few dollars more.

Well I was wrong, and right I guess. It doesn't seem to be the specific card as I'm now using an EVGA gt220 instead of the galaxy gt240 and the same problem persist although I see no problems with the biostar gt9400. Bios settings haven't helped thus far. I'm currently trying an older proprietary driver to see if the problem replicates. We'll see.

Ok, apparently I forgot I was running a beta driver. 195.30 to be exact. It seems that the gt2xx's and that driver do not like each other. However the current stable driver, 190.53, is just fine with the gt220.

Just in case the driver is not the issue -

Something did just jog my memory though. Before Slackware64, I had to adjust vmalloc on 32bit kernels on some systems. You can check your vmalloc with
Look at the VmallocTotal and Used values. I recall having an issue similar to yours, and increasing the vmalloc to much higher number resolved this. On 64bit kernels this is not an issue.
As you can see, I'm using ~312MB of vmalloc, 32bit kernels only allot something like 128MB of vmalloc. Of course if you're using Slackware64