I've recently built a system based on a P6T SE motherboard that has
severe stability issues.
A boot-time (i.e. not in DOS or windows) memory test reports many
errors. The system can sometimes get into Windows but not always.
If hyperthreading is on (in BIOS) the system will spontaneously reboot
within a few minutes of power on. If hyperthreading is disabled then I
will get a BSOD at some point in the first few minutes if in windows.
What I've tried -
Relaxing memory timings
Upping memory as far as 1.66 V
Swapping ALL components (including the motherboard itself) but NOT the
CPU for know good i.e. they work in another system - P6T (not SE
though).
Nothing makes any difference.
My question is - How likely is it the CPU itself is the problem?
Any advice appreciated.
(BTW I've built many systems and never had this much pain before!)
GrahamU wrote:
> I've recently built a system based on a P6T SE motherboard that has
> severe stability issues.
> A boot-time (i.e. not in DOS or windows) memory test reports many
> errors. The system can sometimes get into Windows but not always.
> If hyperthreading is on (in BIOS) the system will spontaneously reboot
> within a few minutes of power on. If hyperthreading is disabled then I
> will get a BSOD at some point in the first few minutes if in windows.
>
> What I've tried -
> Relaxing memory timings
> Upping memory as far as 1.66 V
> Swapping ALL components (including the motherboard itself) but NOT the
> CPU for know good i.e. they work in another system - P6T (not SE
> though).
>
> Nothing makes any difference.
> My question is - How likely is it the CPU itself is the problem?
>
> Any advice appreciated.
> (BTW I've built many systems and never had this much pain before!)
>
You claim "Swapping ALL components" but you don't show that you've changed
the memory. Since your errors seem to indicate memory errors, why not set
the motherboard back to "safe defaults" and stick some different memory in
it -- it doesn't need to be a huge amount, just enough to boot into the OS
and test for a while.
Normally I wouldn't suspect memory as the first culprit on a new build but
after my recent situation (three RMAs and a couple of months of delay time
and I still only have two sticks that work properly) with four 1gB sticks
of Crucial Ballistix memory and an entire menagerie of weird intermittent
seemingly-conflicting symptoms I'm now inclined to suspect memory first
when weird things start to happen.
Re RAM -
I've tried each of the 3 (Corsair) sticks bought for this system
individually and the 3 sticks (different make - Geil - but same speed)
from the other good system - all gave the same unstable results.
Also swapped PSU, Motherboard, graphics card, no change.
On Fri, 06 Nov 2009 10:35:33 -0500, John McGaw <Nobody@Nowh.ere>
wrote:
>GrahamU wrote:
>> I've recently built a system based on a P6T SE motherboard that has
>> severe stability issues.
>> A boot-time (i.e. not in DOS or windows) memory test reports many
>> errors. The system can sometimes get into Windows but not always.
>> If hyperthreading is on (in BIOS) the system will spontaneously reboot
>> within a few minutes of power on. If hyperthreading is disabled then I
>> will get a BSOD at some point in the first few minutes if in windows.
>>
>> What I've tried -
>> Relaxing memory timings
>> Upping memory as far as 1.66 V
>> Swapping ALL components (including the motherboard itself) but NOT the
>> CPU for know good i.e. they work in another system - P6T (not SE
>> though).
>>
>> Nothing makes any difference.
>> My question is - How likely is it the CPU itself is the problem?
>>
>> Any advice appreciated.
>> (BTW I've built many systems and never had this much pain before!)
>>
>
>You claim "Swapping ALL components" but you don't show that you've changed
>the memory. Since your errors seem to indicate memory errors, why not set
>the motherboard back to "safe defaults" and stick some different memory in
>it -- it doesn't need to be a huge amount, just enough to boot into the OS
>and test for a while.
>
>Normally I wouldn't suspect memory as the first culprit on a new build but
>after my recent situation (three RMAs and a couple of months of delay time
>and I still only have two sticks that work properly) with four 1gB sticks
>of Crucial Ballistix memory and an entire menagerie of weird intermittent
>seemingly-conflicting symptoms I'm now inclined to suspect memory first
>when weird things start to happen.
GrahamU wrote:
> Re RAM -
> I've tried each of the 3 (Corsair) sticks bought for this system
> individually and the 3 sticks (different make - Geil - but same speed)
> from the other good system - all gave the same unstable results.
>
> Also swapped PSU, Motherboard, graphics card, no change.
>
snip...
Then at this point that the only commonality left would be the CPU. Or
maybe some cables. Or maybe even the case shorting something out but
usually when that happens nothing works at all. Any way you can borrow a
identical/similar CPU for a while?
That's my next step. Over the weekend I will swap the CPUs in the good
and bad systems. They are not identical though - the 'good' one is a
i7 940 (Stepping 4 Revision C0/C1) and the 'bad' one is a Intel Core
i7 920 D0 Stepping (SLBEJ).
On Fri, 06 Nov 2009 13:38:27 -0500, John McGaw <Nobody@Nowh.ere>
wrote:
>GrahamU wrote:
>> Re RAM -
>> I've tried each of the 3 (Corsair) sticks bought for this system
>> individually and the 3 sticks (different make - Geil - but same speed)
>> from the other good system - all gave the same unstable results.
>>
>> Also swapped PSU, Motherboard, graphics card, no change.
>>
>snip...
>
>Then at this point that the only commonality left would be the CPU. Or
>maybe some cables. Or maybe even the case shorting something out but
>usually when that happens nothing works at all. Any way you can borrow a
>identical/similar CPU for a while?
On Fri, 06 Nov 2009 14:53:14 +0000, GrahamU
<GrahamUnd_NoSpamHere@supanet.com> wrote:
>I've recently built a system based on a P6T SE motherboard that has
>severe stability issues.
>A boot-time (i.e. not in DOS or windows) memory test reports many
>errors. The system can sometimes get into Windows but not always.
>If hyperthreading is on (in BIOS) the system will spontaneously reboot
>within a few minutes of power on. If hyperthreading is disabled then I
>will get a BSOD at some point in the first few minutes if in windows.
>
>What I've tried -
>Relaxing memory timings
>Upping memory as far as 1.66 V
>Swapping ALL components (including the motherboard itself) but NOT the
>CPU for know good i.e. they work in another system - P6T (not SE
>though).
>
>Nothing makes any difference.
>My question is - How likely is it the CPU itself is the problem?
>
>Any advice appreciated.
>(BTW I've built many systems and never had this much pain before!)
When you swapped the board did you use another of the same
model? I'm wondering if the bios has some bugs and might
eventually be stable with a newer bios version.
If you were using the same windows installation it may be
time to abandon it, may be quite corrupt from the memory
errors during installation or subsequent boots, but it is
curious that you could manage to install windows at all with
the system so instable, leading me to wonder what has
changed between then and now.
It certainly isn't impossible for a CPU to be defective or
damaged from ESD during handling, but otherwise it is one of
the less likely suspects... yet as you rule everything else
out it rises in probability of being the problem. I'd also
recheck the heatsink in case something about it or the board
prevents it seating properly on the CPU, and leave the
system sitting in the bios hardware/health screen to monitor
voltages and temperatures.
I wouldn't bother trying to run or even reinstall
windows/other OS anymore until it can consistently POST and
run memtest86+ for several hours w/o errors. If all else
fails and you are still within the seller warranty return
period for the CPU, I'd go ahead and seek the RMA so you
aren't left with the more lengthly wait on Intel replacing
it.
Unstable new build - P6T SE - Why? 
Linear Mode
