Having some problems with my system lately I've been trying to diagnose. System specs are:
- Foxconn Destroyer 780a based motherboard
- AMD Phenom 9950 overclocked to 3.12Ghz @ 240 bus speed w/1.425 Vcore and cooled by Vapochill LS
- 4GB Crucial Ballistix tracer running at DDR 960 speeds with timings as tight as stably possible @ stock voltage (2.0V)
- 2x GTX 460 in SLI @ 815/2100
- Enermax galaxy 1KW power supply
- 4x Seagate barracuda 500GB HDD's in RAID 0 for storage/gaming array
- Western digital 80GB raptor for system drive
- Auzentech X-meridian
Now for the problem. The system has been running great at these settings for around 8 - 9 months now. Let me also say that before this problem hit, I actually started off with 6GB of RAM. I had two 2GB sticks in channel 1 and two 1GB sticks in channel 2. Out of nowhere seemingly, the system became unstable. I would come home sometimes to an idling system and other times I would come home to a POST screen due to a BSOD. It gradually got worse to the point that I could not have the system turned on for more than a few minutes at a time before it would BSOD. I began tinkering around with parts and discovered that by removing the two 1GB sticks of RAM, it cured the problem. Thinking it to be a bad set of RAM, I carried on...
It ran fine for about a week, then yesterxday it started doing this again. So far, I have gotten a total of 4 different bluescreen, and I've noticed that both an Nvidia driver and win32k.sys are cited. The BSOD's I've gotten are:
IRQL_NOT_LESS_OR_EQUAL
PAGEFAULT_IN_NONPAGED_AREA
MEMORY_MANAGEMENT
SYSTEM_SERVICE_EXCEPTION
There seems to be no rhyme or reason to any specific acticity that triggers each one. They just pop up in a random order. One thing I have noticed, Is that if you turn the system OFF and let it rest/cool down, you can run it for about an hour reliably before it starts crashing again. The longer you run it, however, the less stable it becomes until it get's to the point where it cannot even boot into windows or starts locking up at the POST screen, forcing you to shut it down and let it rest for awhile. After boot-up, I ran Prime95 and it did OK for awhile, but then eventually the workers started erroring out one by one and eventually crashed to BSOD.
I tried restoring my BIOS defaults to 100% stock settings, which actually bought me a slightly longer amount of time before the system becomes unstable but the problem still persists.
Now, here is the testing I have done and I believe I have isolated the problem. I started by removing both video cards and running the onboard video. The problem did not go away. Given the fact that all voltages looked good in the BIOS, I ruled out the power supply and video cards at this point. The next test I did was to remove one of the 2GB sticks of RAM, leaving me running off of only one 2GB stick. I fired up Prime95 and the system became stable at this point. I then removed this stick of RAM, and put the other 2GB stick I initially removed in It's place. The system was once again Prime95 stable. I then re-installed both sticks side by side and I couldn't even get windows to boot again.
At this point, I BELIEVE the memory controller on my CPU has gone belly-up, and I'm looking for the opinions of others to confirm or deny this suspicion.
- Foxconn Destroyer 780a based motherboard
- AMD Phenom 9950 overclocked to 3.12Ghz @ 240 bus speed w/1.425 Vcore and cooled by Vapochill LS
- 4GB Crucial Ballistix tracer running at DDR 960 speeds with timings as tight as stably possible @ stock voltage (2.0V)
- 2x GTX 460 in SLI @ 815/2100
- Enermax galaxy 1KW power supply
- 4x Seagate barracuda 500GB HDD's in RAID 0 for storage/gaming array
- Western digital 80GB raptor for system drive
- Auzentech X-meridian
Now for the problem. The system has been running great at these settings for around 8 - 9 months now. Let me also say that before this problem hit, I actually started off with 6GB of RAM. I had two 2GB sticks in channel 1 and two 1GB sticks in channel 2. Out of nowhere seemingly, the system became unstable. I would come home sometimes to an idling system and other times I would come home to a POST screen due to a BSOD. It gradually got worse to the point that I could not have the system turned on for more than a few minutes at a time before it would BSOD. I began tinkering around with parts and discovered that by removing the two 1GB sticks of RAM, it cured the problem. Thinking it to be a bad set of RAM, I carried on...
It ran fine for about a week, then yesterxday it started doing this again. So far, I have gotten a total of 4 different bluescreen, and I've noticed that both an Nvidia driver and win32k.sys are cited. The BSOD's I've gotten are:
IRQL_NOT_LESS_OR_EQUAL
PAGEFAULT_IN_NONPAGED_AREA
MEMORY_MANAGEMENT
SYSTEM_SERVICE_EXCEPTION
There seems to be no rhyme or reason to any specific acticity that triggers each one. They just pop up in a random order. One thing I have noticed, Is that if you turn the system OFF and let it rest/cool down, you can run it for about an hour reliably before it starts crashing again. The longer you run it, however, the less stable it becomes until it get's to the point where it cannot even boot into windows or starts locking up at the POST screen, forcing you to shut it down and let it rest for awhile. After boot-up, I ran Prime95 and it did OK for awhile, but then eventually the workers started erroring out one by one and eventually crashed to BSOD.
I tried restoring my BIOS defaults to 100% stock settings, which actually bought me a slightly longer amount of time before the system becomes unstable but the problem still persists.
Now, here is the testing I have done and I believe I have isolated the problem. I started by removing both video cards and running the onboard video. The problem did not go away. Given the fact that all voltages looked good in the BIOS, I ruled out the power supply and video cards at this point. The next test I did was to remove one of the 2GB sticks of RAM, leaving me running off of only one 2GB stick. I fired up Prime95 and the system became stable at this point. I then removed this stick of RAM, and put the other 2GB stick I initially removed in It's place. The system was once again Prime95 stable. I then re-installed both sticks side by side and I couldn't even get windows to boot again.
At this point, I BELIEVE the memory controller on my CPU has gone belly-up, and I'm looking for the opinions of others to confirm or deny this suspicion.