What would cause this loss of performance and glitching?

Lately I've been losing performance in games and experiencing strange glitches. I know its a HW problem:

I think it might even be the CPU because. I just noticed under load my 6550 hits 80c. that's hot.
I'm on stock cooling with the overclock in my sig.
It could also be the Mobo or GPU.. Please help me.
This issue began as visibly reduced framerates in my games that I play. At first, it only started to happen when I exited one game and loaded another-- as if some kind of hardware resource wasn't freed up. If I restarted the machine the problem would go away. Now restarting the machine has no effect.

Then I started seeing weird video glitches like these:



You can see in the first two that the glitch is the same on the two different maps. The glitch pattern only changes when I restart the game.










I've confirmed this is a hardware problem because I've had these same issues on two completely seperate drives with seperate OS's which do not see eachother.

They even have totally different:

Motherboard drivers

Videocard drivers

Sound drivers

just to be sure.

I also get very uneven performance. For example last night the World in conflict benchmark got me an average rate of 48 with a minimum of 25. Today it was 48 and 31. How can performance vary so much?

I ran MEMTEST in safemode to test the most ram possible and completed 2000% (15HRS). 0 errors on both sticks.

My system is:
Asus P5K Premium P35
BFG 8800GTS g92
ASUS Xonar D2 PCI
2x1024 Corsair Dominator 1066
5-5-5-15-2T 518Mhz (at POST says 1039)
E6550@433x7 3.04Ghz 1.33ish Vcore
Antec Sonata w/500W earthwatts

all voltages auto. Ram set to 2.1V as per mfg. spec.

HARDWARE FAILURES SO FAR:

Seagate HDD Failed 5 months ago. Replaced with WD
(bizzarely- my laptop, on AC power, also had the HD die a week later)

RAM (mushkin
1066) failed 2 months ago. Replaced with corsair.

BAD STUFF THAT HAS HAPPENED TO SYSTEM:

1Kinda got hot at one point. Forgot to clean it for awhile. When Mushkin ram failed opened case and dust was completely covering the CPU HS resulting in crappy cooling. Did not check temps at that time but when everything is clean my peak under load is about 50 celsius on the CPU and 60-65 on the GPU

2 I smoke a lot. System is in a smokey room. Smoke=tar=dust STICKS to stuff... I don't know if this tar conducts electricity or not... Just putting it out there to help you help me

3 I once removed the HS from the GPU (about six months ago. It has worked fine since AFAIK. Did not begin getting these problems until 1 month ago)

After I blew all the dust out I replaced the HS properly afaik. No thermal pads were damaged AFAIK.

artifacting that starts seemingly from nowhere has been, in my experience, the sign of a dying card. That said, I'd check on a few things before you start worrying about buying a new one, or even borrowing one to verify. First, is the card overclocked? It might just be running too hard and causing the artifacts. Second, do you have anything like rivatuner running to monitor your card's temps? If not, I'd strongly recommend installing something and take a look at your temps under a heavy load. Since you did remove the heatsink at one point, it might not have gotten reinstalled properly. If it's been running for six months at overly high temperatures, you may have been giving it a slow, painful death. EDIT: okay, just noticed your video card temps. a little high, but probably not doing any damage.

If the card isn't overclocked, and your temps check out okay, I'd start looking for a card to test in it's place if possible. If you don't know anyone who can lend you a card for a couple hours, you might have to just replace the card. While 80 degrees on your cpu is hot, that's probably not your culprit in this case. You could drop your overclock to test it, but my guess is that you won't find the problem fixed.

overclcoedk too much.

Clean out your CPU and GPU Heatsink and fan. Verify the GPU fan is spinning and not warbling any.

Try running 3dmark, see if it gives you a problem.

If your cpu is hitting 80C at load, you have a problem there. Check to make sure that all the pushpins on the stock cpu cooler are indeed fastened down tightly and correctly. Like PhysX mentioned, maybe try backing off your OC to see if that helps.