I'm about to lose my head over this one.
My system specs:
Asus p4t533 motherboard (new one with fan on northbridge)
P4 2.66 mhz
1 x 512 samsung rim-4200 32-bit rdram (everything is default - FSB is running at 533 mhz, the voltages are set to auto, etc.)
geforce3 (the first kind - not a "ti")
creative audigy2 OEM
NEC dvd-rom drive
adaptec scsi adapter w/ 8x plextor cdrw
maxtor 60 gig ATA133 drive
ibm deskstar 60 gig ATA100 hard drive
maxtor 30 gig ATA 133 drive
antec true power 430 watt p/s
3com network adapter
Now for the problem:
Well when I first set this thing up, WinXP installed fine (I had the 2 60 gigs on a RAID 0 stripe with the onboard RAID, and I did not have the soundcard) and seemed to be running ok until I started installing and running various games (medal of honor, simcity 4, warcraft 3, etc...) and then after about half an hour running the games they would suddenly quit out to the desktop or throw an exception 0xC0000005. Restarting the game did little good as it would always crash in just a few minutes of play after the first time. I thought "ok must be the memory or the video card" so I ran prime95 and sure enough it failed in a few seconds. (by the way the RMA on the ram w/ googlegear had already expired and I couldn't find anything on samsung's website about RMA)
I then run memtest86 from boot for a very long time (hours and hours) and I get absolutely no errors. I do this several times never getting an error, but prime95 consistenty quits during the first test. I am not overclocking at all, and I have 5 fans on my case and the CPU never gets above 48 celcius, so I sort of ruled out heating.
I then switch out my Geforce3 for a spare Geforce2 GTS I had laying around. The same thing still happens. (Mind you I have reformatted countless times with both Winxp and win2k by this time hoping desperately the problem would go away).
I then think "maybe the RAID controller is screwed up" so I disable the raid and set it all up on the regular IDE channels.
The system seems to run a bit better, but not really and it still crashes out of games (and some other programs like IE) consistently.
I know for sure the sound card is not the issue as I just added it recently and the errors are the same.
The case does not get hot and it is very well ventilated.
(using the big silver antec case with the air filter on the front).
I have one fan on the front blowing in and 2 fans on the back and one on the side blowing out.
Since memtest86 gives up no errors after many many hours of running (probably about 10), I feel like it is safe to assume there is no problem with the CPU, RAM, or motherboard (as the website for memtest86 says that all of these components should be assumed stable if there are no errors). Am I in error for assuming this?
Some hypotheses:
Could it be the I/O subsystem? Could this somehow "interfere" with the system memory to cause spontanious errors? (I have run HD benchmarks that check out fine)
Could it be the SCSI card? Is it at all possible it could be screwing everything up? It's such a hassle to remove it I don't really want to bother unless it is true it could be messing things up (but I certainly will if it's necessary).
Could the memory itself be overheating, but not the CPU, causing the errors? I don't really have much experience with heating issues, but since once the programs crash they tend to crash alot immediately afterwards it seems the memory could be overheating. If it is, how does one cool just the memory, as the rest of the computer is running at normal temps? (I have reseated it a few times to try to fix it.)
Any help/suggestion is GREATLY appreciated, I am very thankful for all of you on these forums that are so willing to help.
-Thorstadt
My system specs:
Asus p4t533 motherboard (new one with fan on northbridge)
P4 2.66 mhz
1 x 512 samsung rim-4200 32-bit rdram (everything is default - FSB is running at 533 mhz, the voltages are set to auto, etc.)
geforce3 (the first kind - not a "ti")
creative audigy2 OEM
NEC dvd-rom drive
adaptec scsi adapter w/ 8x plextor cdrw
maxtor 60 gig ATA133 drive
ibm deskstar 60 gig ATA100 hard drive
maxtor 30 gig ATA 133 drive
antec true power 430 watt p/s
3com network adapter
Now for the problem:
Well when I first set this thing up, WinXP installed fine (I had the 2 60 gigs on a RAID 0 stripe with the onboard RAID, and I did not have the soundcard) and seemed to be running ok until I started installing and running various games (medal of honor, simcity 4, warcraft 3, etc...) and then after about half an hour running the games they would suddenly quit out to the desktop or throw an exception 0xC0000005. Restarting the game did little good as it would always crash in just a few minutes of play after the first time. I thought "ok must be the memory or the video card" so I ran prime95 and sure enough it failed in a few seconds. (by the way the RMA on the ram w/ googlegear had already expired and I couldn't find anything on samsung's website about RMA)
I then run memtest86 from boot for a very long time (hours and hours) and I get absolutely no errors. I do this several times never getting an error, but prime95 consistenty quits during the first test. I am not overclocking at all, and I have 5 fans on my case and the CPU never gets above 48 celcius, so I sort of ruled out heating.
I then switch out my Geforce3 for a spare Geforce2 GTS I had laying around. The same thing still happens. (Mind you I have reformatted countless times with both Winxp and win2k by this time hoping desperately the problem would go away).
I then think "maybe the RAID controller is screwed up" so I disable the raid and set it all up on the regular IDE channels.
The system seems to run a bit better, but not really and it still crashes out of games (and some other programs like IE) consistently.
I know for sure the sound card is not the issue as I just added it recently and the errors are the same.
The case does not get hot and it is very well ventilated.
(using the big silver antec case with the air filter on the front).
I have one fan on the front blowing in and 2 fans on the back and one on the side blowing out.
Since memtest86 gives up no errors after many many hours of running (probably about 10), I feel like it is safe to assume there is no problem with the CPU, RAM, or motherboard (as the website for memtest86 says that all of these components should be assumed stable if there are no errors). Am I in error for assuming this?
Some hypotheses:
Could it be the I/O subsystem? Could this somehow "interfere" with the system memory to cause spontanious errors? (I have run HD benchmarks that check out fine)
Could it be the SCSI card? Is it at all possible it could be screwing everything up? It's such a hassle to remove it I don't really want to bother unless it is true it could be messing things up (but I certainly will if it's necessary).
Could the memory itself be overheating, but not the CPU, causing the errors? I don't really have much experience with heating issues, but since once the programs crash they tend to crash alot immediately afterwards it seems the memory could be overheating. If it is, how does one cool just the memory, as the rest of the computer is running at normal temps? (I have reseated it a few times to try to fix it.)
Any help/suggestion is GREATLY appreciated, I am very thankful for all of you on these forums that are so willing to help.
-Thorstadt