my computer's been really sluggish the past few weeks when running games. i'll get incredibly low FPS and sometimes it just crashes all together. i assumed this was the result of some virus i had picked up, so i reformatted and reinstalled windows.... but the problems remian... i mean i'm getting like 10 fps on old games like BF 1942. i have a 3.2ghz p4 2 gigs of pc3200 nvidia 6800 and a freaking raid 0 running this game...

it must be hardware.. i ran memtest 86 with no errors till the 8th hour of testing in which it got 2 errors in a row very quickly.

could the memory be the only problem? because my problem is constant and it took memtest 8 hours to find any.

any other diagnostic tools you know of i'd like to try i suppose.


[Edited on November 28, 2006 at 7:23 PM. Reason : .]

11/28/2006 7:22:58 PM

prime95

11/28/2006 8:45:04 PM

thats just gonna tax the hell out of the CPU...its not really diagnostic software.

11/28/2006 9:04:43 PM

got any parts you can swap out with to start eliminating things?

11/28/2006 9:08:20 PM

Sounds like a graphics card issue. Bad RAM could cause this too.

[Edited on November 28, 2006 at 10:01 PM. Reason : ]

11/28/2006 10:00:59 PM

it will tell you real quick if something is wrong

11/28/2006 10:10:13 PM

maybe ur PSU isnt giving enough power to properly run everything

11/28/2006 10:26:32 PM

check your video card and CPU fans and make sure they're working

if that's a Prescott 3.2 it runs hot as a motherfucker and can very easily encounter a condition where the processor will underclock itself to maintain a safe operating temperature, I'd also make sure the heatsink was clean, free of airflow obstructions, etc

[Edited on November 28, 2006 at 10:58 PM. Reason : ~]

11/28/2006 10:55:12 PM

my temps are ok on cpu and video card, and my voltages are good on the video card... i guess that means the PSU is ok. i've switched out the memory using only 1 gig at a time and both behave the same...

my other desktop has a nvidia 7800, but it's pci-e x16 and this 6800 is AGP so i can't switch it out....

could the psu go bad without affecting the video card voltages very much?

11/28/2006 10:59:30 PM

the PSU voltages might only be dipping under load, but generally with insufficient voltage the result is logic failure / data corruption, and not so much the freaky shit you're seeing

[Edited on November 28, 2006 at 11:05 PM. Reason : clarification]

11/28/2006 11:01:20 PM

What graphic settings are you trying to run these games on? Will lowering it help any?

11/29/2006 3:47:42 AM

could you switch out the power supply for troubleshooting?

can you get ahold of another video card to use for testing? who mkaes your video card?

11/29/2006 11:01:00 AM

yea i can try another psu from a computer my roomate is using, it's a really nice 500w one. the one i'm using is a thermaltake 420w i think.

the video card i'm using is an asus nvidia 6800. and no i dont have one to switch out.

but what really is weird is i'm averaging the same crappy fps with really LOW settings (640x480 and every thing turned down or off) as i am with fairly high settings (1280x1024 max view distance every thing up)

like i can play CS:source with high quality settings at the same framrate as BF1942 with low settings.. it's weird as hell... that's why i origionally thought it was some sort of virus/malware.

11/29/2006 11:47:23 AM

Quote :

"if that's a Prescott 3.2 it runs hot as a motherfucker"

is that true for the celeron d's as well? because my processor nearly cooked itself before i exchanged it for another one, which has been running fine for months

11/29/2006 12:04:45 PM

well i tried playing oblivion with the same result, except after about 10 minutes my hands start looking like edward scissorhands and graphic spikes start coming out all over the characters body lol....

so i guess it's the vid card.. i'm downloading 3Dmark06 to try to kill it and see if that works.

11/29/2006 4:40:27 PM

^^ not really sure what the Celeron D is core-wise but I'm pretty sure that all x86 Intel chips after the P-III series have featured the thermal scaling shiznizzle, and P-III's had thermal protection as well but I believe it was a shutoff.

^ still sounds a lot like a thermal problem, did you check the fan/blower on the video card? are there any PCI cards in the slot immediately neighboring your video card? how/when are you checking the temperatures? the card's idle temperature won't really indicate anything, and it will cool to within several degrees of its idle temperature fairly rapidly after the load has terminated. you need to see it's temperature under load, if you aren't try running one of your games in a windowed mode and watch the temperature.

11/29/2006 5:14:47 PM

UPDATE: apparently the smartdoctor program that governs fan speed, temps and voltages of the card for reset to defaults and the fan wasn't going full speed. i fixed it and the graphic problems are much better, no more lag.... but games still crash probably once every 30 mins on avg..... no idea why... and now that i have smartdoctor set up correctly it's showing at random times the Vcore will spike up halv a volt, or sometimes go to 0 volts for a couple seconds..... this usually happens when i'm alt tabbing... so it may just be a software thing with smartdoctor.

11/30/2006 8:17:32 PM

well, if your Vcore was actually at 0 you wouldn't be reading that value off the screen, and probably the same for spiking half a volt with the addition of needing a new CPU after the first occurance... odds are that's just a noisy reading

now that the graphics problem is (hopefully) out of the way, try running a CPU-specific test (e.g. prime95 mentioned above) and set it for small or in-place to stress only the CPU and not the memory subsystem.

could also be related to that memtest86 failure, I don't recall if memtest86 has an option to run a specific test repeatedly but if it does you should run just the failing test pattern in a loop to see if it's a specific pattern of access that's leading to the errors.

12/1/2006 5:10:43 PM