Sandy 2500K Or Asus P8P67 Pro looks to be dying

SimonMaltby · 8 Jan 2011 at 22:39

Well I have had the board and chip for 2 days, yesterday it ran prime at 4.6GHz and would superPi at 4.7GHz

For the last few hours I have been getting hardware blue screens in windows even running everything at stock

I can rule out RAM because I have 2 sets.
I have reseated the CPU several times, but temps seem fine.
I have flashed backwards and forwards between the three Asus bioses
I have been running the chip at between 1.3 and 1.35v with a couple to test runs at 1.4v

I think the CPU is dying which would be very bad news.

Going to let everything cool down, reseat the CPU with the supplied Heatsink and try again, but I'm not very hopeful... this has all the signs of a chip in meltdown.

RJC · 8 Jan 2011 at 22:49

What PSU are you using, had a faulty one give me grief before.

SimonMaltby · 8 Jan 2011 at 22:50

RJC said:
What PSU are you using, had a faulty one give me grief before.

Brand new OCZ 700W.
It does not feel like PSU the voltage lines reported in the bios look fine and stable.

JonnyGLC · 8 Jan 2011 at 22:52

How high have you had the CPU? Both clock and voltage wise

tocasa · 8 Jan 2011 at 22:56

Have you got the ASUS software loaded?
I had a problem with mine, when I clicked on some of the eco options. I can't remember exactly what it was called, because I have uninstalled it.

This was undervolting my CPU to save energy, causing blue screens. But that was only when overclocked, it seemed ok at stock. Took me a few reboots to figure out though.

tocasa · 8 Jan 2011 at 22:57

tocasa said:
Have you got the ASUS software loaded?

Or the EPU switch might have the same effect - is it on?

SimonMaltby · 8 Jan 2011 at 23:19

A random name said:
How high have you had the CPU? Both clock and voltage wise

CPU Max it has managed to get into windows at (i.e. once the Turbo cuts in on boot) is about 4.75 ans the max I have set the vcore is 1.4v in the bios

tocasa said:
Or the EPU switch might have the same effect - is it on?

Yea I had an issue with that earlier and noticed the vcore and vdimm were lower than i set them. But thats all switched off now.

UPDATE:
Switched the CPU cooler from H50 to stock Intel
Have re-cleared CMOS
But no good. I get a blue screen with "an uncorrectable hardware error" and thats with everything at stock.

The only 2 other things I can try as I have spares is the GPU and PSU.

chippy08 · 8 Jan 2011 at 23:24

Dont sound good new parts dying already will watch post with interest before i buy.

mrbios · 8 Jan 2011 at 23:33

You able to take a photo of the blue screen so we can see exactly what's coming up?

Also, i know it says hardware error, but is there any possibility of windows file corruption caused by your overclock during an unstable moment? (just trying to rule out everything)

SimonMaltby · 8 Jan 2011 at 23:45

mrbios said:
You able to take a photo of the blue screen so we can see exactly what's coming up?

Will Do.

mrbios said:
Also, i know it says hardware error, but is there any possibility of windows file corruption caused by your overclock during an unstable moment? (just trying to rule out everything)

No chance. I thought the same which is why I set everything back to stock and am now just trying to install windows anew. I get the blue screen once all the files are loaded and the install re boots for the first time.

PSU and GPU both tried with no difference.

I have now turned off Turbo, set the CPU to one core and disabled as many motherboard addons as I can. Its just installing now. Will post back with either the blue screen shot or result in a few minutes

SimonMaltby · 8 Jan 2011 at 23:59

I think I have exhausted everything I can. It must be either the CPU or the Motherboard.

QuackPot · 9 Jan 2011 at 00:15

Don't overclock.

/thread.

SimonMaltby · 9 Jan 2011 at 00:22

QuackPot said:
Don't overclock.

LOL

VaderDSL · 9 Jan 2011 at 00:27

Can you rule out the Ram though? maybe try a quick memtest?

mrbios · 9 Jan 2011 at 00:28

I'd say try running memtest just as a precaution, even though you have two sets of memory. Beyond that, considering all the things you've tried i think it's time to hit the RMA train

EDIT: VaderDSL is too quick for me!

SimonMaltby · 9 Jan 2011 at 00:38

mrbios said:
I'd say try running memtest just as a precaution, even though you have two sets of memory. Beyond that, considering all the things you've tried i think it's time to hit the RMA train

EDIT: VaderDSL is too quick for me!

Memtest does not work with Sandy Bridge yet.
Fount 4.2 Beta that does. Will try now.

VaderDSL · 9 Jan 2011 at 00:40

4.10 has preliminary support for sandy bridge, might be worth giving that a shot?

http://www.memtest.org/#history

SimonMaltby · 9 Jan 2011 at 00:47

VaderDSL said:
4.10 has preliminary support for sandy bridge, might be worth giving that a shot?

4.2 beta is running now..... half way through test 6 with no issues so far

VaderDSL · 9 Jan 2011 at 00:49

If the RAM is fine and the test run ok, perhaps the HDD is on the way out?

SimonMaltby · 9 Jan 2011 at 00:52

VaderDSL said:
If the RAM is fine and the test run ok, perhaps the HDD is on the way out?

Just eliminated that. I swapped the OCZ SSD for a std SATA drive and the same things happens.

The best result i have got is by turning off Turbo and setting the multi to it's lowest - 16x This installs windows and only fails on the very last step... with the same error.

This leads me more and more to the CPU... because if I can get it to go a bit further by slowing it down there must be a correlation.

Once Mem test completes and assuming there is no error, i'm going to assume this is the CPU and order a new one (probably a 2600k instead) then assuming that works I will RMA this one. If the new CPU still fails then I know its the motherboard.