I think my CPU may be dieing...

I've always been afraid of something like this happening, and I'd really hate to say this, but I think my CPU is seeing out its last week or so of its life

Syntoms
I recently sold my Gigabyte K8NS Ultra that I ran 275x11 - 1.55vcore for about 3 months straight. It ran 110% stable with extremely nice temp sofr about 21 ~ 29C. It never crashed on me, once.

Upon selling that I bought me a DFI Ultra-D (Might I add I love this board). After about 2 or 3 weeks of learning the board and its numerous settings, I finally came up with a pretty decent OC, 270x10 - 1.584vcore. T'was prime stable 27+ hours. Pretty dcent temps for air; 24 ~ 32C.

Recently, Day of Defeat Source was released and I'm addicted to it. 2 nights ago I played it for roughly 7 hours straight of full GPU & CPU load - not a hitch insite. So i went to sleep....woke up 6 hours later, A friend wanted me to play a quick game. I join up and after about 15 minutes of playing I got a hit with a BSOD stick! Kinda worried me...but I thought nothing of it and a good restart couldnt hurt. Got in game...10 minutes later..WACK!1 BSOD!. Now I was worried. Using the same timings and voltages, I backed off to 260x10, which is OCCT, Prime95 & PI stable for 48+ hours. BSOD in about 10 minutes

I imediantly shot off my computer, reset the CMOS for 5 minutes, boot back up to stock everything, 200x11. Ran Memtest86 to see if my ram was crapping out, 2 passes of all the tests cleared with flying colors. Got myself in for a quick game, BSOD Prime was failing in 1 minute...Random Windows & crashes randomly throughout there.

Is my CPU dieing on me?

ETC Info.

Rails (Measured with multimeter)

12+
Idle: 12.09v
Load: 12.01v

5+
Idle: 5.12
Load: 5.04

What did you expect running a 30%+ overvolt?

Well if you've really been running 1.854 thru it thats quite possible....

If you've got the time man, stick it back in the gigabyte that ninja has now with the same stix and PSU and see what happens, could be a bad board also, or maybe it doesnt like your PSU??

If you really pumped 1.854 thru it that may be the problem.

Ok, now that I have the snarky comment out of my system, here is some real advice.

Back EVERYTHING back to stock voltage and frequency. See if the problems persist. Next start playing with one thing at a time to see what causes problems. (i.e. increase cpu mhz, leaving memory at stock and vice versa.) Try running with different ram if you can. If you are lucky, your computer will continue to run at stock or *slightly* overclocked speeds for some time to come. But I think your heavy overclocking days are over with that particular cpu and/or ram.

Also, double check that all your cooling is properly working.

Oh holy ****s on TYPO

1.584 VCORE

Going to edit my post now.

ziggo0 said:

Oh holy ****s on TYPO

1.584 VCORE

Going to edit my post now.

Click to expand...

LOL

i would try the other board, i bet you got a bum board.
or its a bios issue...

Yeah but remember me telling you on MSN how my bios reading was idling like 38C :S

I think my CPU might actually be dieing

bios always reads 2-4C more than software which may not be accurate to begin with anyway.

i would put it back in the gigabyte and memtest it and prime it and if it passes there's your answer.

Or buy my san diego for ninja and try that one in your DFI

I blame the Rosewill PSU, try another one before trying a new CPU.

yeah its PSU/BOARD related i would think

What are your BSOD errors??? You might have some corrupted windows files, you could try formatting and reinstalling windows.

Why the heck does everyone rag on this PSU? Yeah sure its a generic brand but it owns...the rails are stable as hell, its native 24 pin, what more could you ask for? Its no better than a forton or ocz. If a Forton or OCZ can give strong stable rails, high ampage and decent wattage, there is no difference.

My external temp monitor was reading 29 ~ 30F load whenever it was ****ing up.

My BSOD errors give no reason...not like irq equal to less blah blah blah...just a straight up error. Whenever it 'BSOD' at stock, it does it so fast i cant see it.


I guess Ill grab that board back and try it on there..Personally I dont care if the CPU is dead...its still under warranty.

Disable automatic reboot after errors. Then you can see what the errors are.

Personally Im apt to blame the board and not the CPU

Sentential said:

Personally Im apt to blame the board and not the CPU

Click to expand...

me too
the only think changed was the board so defenitivily the board or bios setting check them then post the result

Yeah...I think your onto something. I think the board is too almost....it has a hard time running anything...even my previous OC's...'ll get that board from him tonight.

I used i Rosewill 550W before this Fortron and my Rosewill couldnt hold stable rails to save its own *** ... so it got replaced obvious reasons now my CPU is stable

The rails might seem stable but you would have to check for voltage ripple with an oscilloscope. I think the latest ATX specs are within 50 mv on the 3.3 V rail and 120 mv on the 12 V rail. Exceeding these specs will lead to instability. Just pointing out that seemingly stable rails does not necessarily indicate a properly functioning PSU. I am all about cost effectiveness, but there is no reason to skimp on the PSU.

ziggo0 said:

Why the heck does everyone rag on this PSU? Yeah sure its a generic brand but it owns...the rails are stable as hell, its native 24 pin, what more could you ask for? Its no better than a forton or ocz. If a Forton or OCZ can give strong stable rails, high ampage and decent wattage, there is no difference.

My external temp monitor was reading 29 ~ 30F load whenever it was ****ing up.

My BSOD errors give no reason...not like irq equal to less blah blah blah...just a straight up error. Whenever it 'BSOD' at stock, it does it so fast i cant see it.


I guess Ill grab that board back and try it on there..Personally I dont care if the CPU is dead...its still under warranty.

Click to expand...

Alright, you are asking for help and when you hear something you don't like you start getting mad?
Fine, its your problem. RMA your CPU and will have the same problems in matter of weeks/months.

RAM or Mobo would be my instinct..I didnt read allthe way down....did you try resetting BIOS and starting from scratch?