I've been troubleshooting my recently upgraded computer for the last several days and think that I have found a problem with one of my G.Skill DIMMs. Here's what is going on.
My system:
ASUS P8Z68-V (0902 bios), i5-25000
2x4GB G.Skill RipjawsX F3-12800CL8D-4GBXM
MSI GTX 560 Ti Twin Frozer II,
AData 64GB SSD,
WD Caviar Blue 1TB
Hitachi 1TB
CoolerMaster Hyper 212+
Corsair AX750 PSU
I noticed instability problems when playing games (DDO mostly) not long after I assembled all the components - bluescreen errors or my video driver would crash. I started off with a Corsair TX650 PSU that made an odd high-pitched noise when games would start or when in certain video modes. I considered it suspect and replaced it first. My box was a lot quieter afterward, but the instability issues remained.
Next I suspected the motherboard because I tried to run memtest86+ and it just locked up, even when using a different set of memory (Corsair Vengeance LP from a local Frys). The motherboard replacement seemed like it fixed the issue for about 2 hours, then the problem started up again.
At this point I figured it had to be memory. I downloaded the latest memtest86+ off memtest.org and this time it worked. I ran it all night long and never found a problem. Next I tried Prime95 in blended mode. With that I can duplicate the problem very quickly.
When running Prime95/blend with just DIMM A installed (in slot A2, which ASUS recommends), my box is stable. I ran Prime95 for several hours and was able to browse the web and play a video at the same time.
When running Prime95/blend with just DIMM B installed (in either slot A2 or B2), I start having problems very quickly. Either a bluescreen or application errors - prime95 workers will return errors and stop, Chrome will show the "aw snap" screen, explorer windows won't open, etc.
Last night I decided to run memtest86+ on individual DIMMs. With DIMM A installed, it ran for 3 hours with no error at all. With DIMM B installed, I ran it all night long. No errors in 17 passes. Definitely not what I expected there.
What do you guys think? Seems like a memory problem to me, but I don't get why memtest86+ can't find an error.
Thanks in advance,
Mike
My system:
ASUS P8Z68-V (0902 bios), i5-25000
2x4GB G.Skill RipjawsX F3-12800CL8D-4GBXM
MSI GTX 560 Ti Twin Frozer II,
AData 64GB SSD,
WD Caviar Blue 1TB
Hitachi 1TB
CoolerMaster Hyper 212+
Corsair AX750 PSU
I noticed instability problems when playing games (DDO mostly) not long after I assembled all the components - bluescreen errors or my video driver would crash. I started off with a Corsair TX650 PSU that made an odd high-pitched noise when games would start or when in certain video modes. I considered it suspect and replaced it first. My box was a lot quieter afterward, but the instability issues remained.
Next I suspected the motherboard because I tried to run memtest86+ and it just locked up, even when using a different set of memory (Corsair Vengeance LP from a local Frys). The motherboard replacement seemed like it fixed the issue for about 2 hours, then the problem started up again.
At this point I figured it had to be memory. I downloaded the latest memtest86+ off memtest.org and this time it worked. I ran it all night long and never found a problem. Next I tried Prime95 in blended mode. With that I can duplicate the problem very quickly.
When running Prime95/blend with just DIMM A installed (in slot A2, which ASUS recommends), my box is stable. I ran Prime95 for several hours and was able to browse the web and play a video at the same time.
When running Prime95/blend with just DIMM B installed (in either slot A2 or B2), I start having problems very quickly. Either a bluescreen or application errors - prime95 workers will return errors and stop, Chrome will show the "aw snap" screen, explorer windows won't open, etc.
Last night I decided to run memtest86+ on individual DIMMs. With DIMM A installed, it ran for 3 hours with no error at all. With DIMM B installed, I ran it all night long. No errors in 17 passes. Definitely not what I expected there.
What do you guys think? Seems like a memory problem to me, but I don't get why memtest86+ can't find an error.
Thanks in advance,
Mike
Comment