New Build Has Stability Issues

[PolygonGTC]

I just upgraded my system, the specs are in my sig. It all started with hard locks. No BSOD, just a hard lock at which point I would have to hold down the power button and restart. It's very random but is consistent when the system is pushed. However, sometimes it will take some time and sometimes it will happen quickly. Now it is getting to the point were it will just reboot. It has done this right after logging into Windows as well. Another new development is that I will get static through the speakers ocassionaly. Also, when messing with the BIOS it would not boot. It would post but Windows failed to load until I went back to fail safe defaults.

I've been trying to run Prime95 and I get this result:

[Sat Jan 16 23:36:04 2010]
FATAL ERROR: Resulting sum was 3.206511881782796e+016, expected: 4.764133057460626e+016
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Resulting sum was -804318095685795, expected: 1.331283465043802e+016
Hardware failure detected, consult stress.txt file.

When it does this it consitantly fails on worker 7 and 8. So, I am assuming that the fourth core is bunk. I've also run memtest86 three days straight on it with no failures. Here are some other solutions/conjecture I've gotten:

1. Reinstall Windows with a slower optical drive.
2. Memory is being undervolted, drive up the voltage. Which it is being undervolted by the auto settings.
3. Could be a northbridge issue.

I would be pretty confident that it's a CPU but the static through the speakers and the failures to boot after changes to the BIOS have me wondering. Also, I am not over clocking.

Any advice?

BTW: OS is Win7 x64 Pro.

 

[PolygonGTC]

Now I just tried running IntelBurn test which I found in a sticky up top. It failed right off the bat with a hard lock and tone through the PC speaker. Booted it back up and it BSODed and the only thing I could really make out before it rebooted was something about a USB driver. Then it came back up again and hard locked with static through the speakers. Then I booted it again and it is working for now.

I doubt this is something wrong with the installation.
I also doubt it's the memory even though it is being undervolted.

However, I don't know if this is a MOBO issue or a CPU. Does anybody have any ideas?

 

[Dangman]

Bump up the voltage for the CPU just a tad and see what happens. Also make sure that the RAM voltage does not exceed more than 1.65V.

 

[PolygonGTC]

Bump up the voltage for the CPU just a tad and see what happens. Also make sure that the RAM voltage does not exceed more than 1.65V.

Even though Mushkin rates it at 1.85-1.95v? It is currently under 1.65v though.

I'll bump up the VCORE and see where that gets me.

 

[lt_shiro]

At that voltage it could very easy kill your cpu. I suggest buying new memory as those Blackline were for old 775 DDR3 boards and not 1156 CPU.

 

[Dangman]

Even though Mushkin rates it at 1.85-1.95v? It is currently under 1.65v though.

If you had the RAM voltage under 1.65V the entire time you were running Memtest, then the RAM should be ok to run at that voltage. With that said, as lt_shiro, any more than 1.65V and you do risk damaging the CPU or causing instability.

Hmmm, now that I think about it, try a different set of RAM that was actually rated at 1.65V and below. Got a hunch.

 

[MJZ]

Check your memory voltage. The board probably set the memory to 1066 @ 1.5V on Auto, but make sure.

 

[PolygonGTC]

Bumped up the VCORE to 1.25 and it's still having all sorts of strange failures. It wouldn't boot to Windows initially and I had perform a check disc because the system volume was corrupted. Even after that it locked again under stress. I've bumped the VCORE up to 1.3 now and I'm testing again.

If it fails I'm going to call Mushkin on Monday about the memory. Just odd that it would have issues even though it passed memtest for three days at that voltage. Then again.

Now, would you guys say that it's safe to rule out the MOBO here?

At that voltage it could very easy kill your cpu. I suggest buying new memory as those Blackline were for old 775 DDR3 boards and not 1156 CPU.

If you had the RAM voltage under 1.65V the entire time you were running Memtest, then the RAM should be ok to run at that voltage. With that said, as lt_shiro, any more than 1.65V and you do risk damaging the CPU or causing instability.

Hmmm, now that I think about it, try a different set of RAM that was actually rated at 1.65V and below. Got a hunch.

Good hell!!!

Thanks for point this out. Now I just realized I ordered the wrong kit. There were two that I was looking at and I went back and I ordered the wrong. There is one specifically made for the P55 chipset and I thought I had ordered that.

ROFL..... Anyone want this set?

Check your memory voltage. The board probably set the memory to 1066 @ 1.5V on Auto, but make sure.

They're set to 1333MHz @ 1.5v on auto right now.