124 BSOD Now Rig Won't Post

[why_sleep]

*Update*

Tried everything short of buying another mobo/cpu to swap them out. The 980x had just worn out before this happened so they can't rma, EVGA won't rma the board either due to the fact that the bsod happened during an "overclock session" (?) so any other suggestions are highly appreciated

Hi OC3D Community,

About 6 months ago I bought a barely used 980X from a friend, with eventual plans for an x58 workstation. Finally got the rig put together recently and everything was running great at stock clocks/voltages for weeks. After weeks of research (much from this forum) I decided to push the chip a little yesterday. It went to 4.0 ghz very easily & with fairly low voltages, passing 50 runs of linx & 12 hours of p95. This morning, while playing with e-leet, I lowered the cpu vtt from 1.2v to 1.075 and received a 124 bluescreen error. I've read that error code means not enough or too much voltage on the vtt or vcore, so thinking I just pushed the vtt somewhere it didnt like I tried to power back on and nothing; the rig won't post bios, nothing. Tried clearing the cmos several times, resetting the battery & checked all ram dimms. Also tried a different PSU, just off chance, no luck. I don't have a bios template to supply for obvious reasons but here's the specs:

EVGA E770 Classified 3

i7 980X @ 4.0ghz

Noctua NH-D14

Generic 250gb Western Digital Sata Drive

GTX 470 @ stock

12gb (4gbx3) Corsair Dominator GT 2000mhz

Cooler Master Silent Power Pro Gold 1200W

& Voltages(what I can remember off the cuff, at least) :

vdroop off

vcore: 1.25v in e-leet, 1.24 in bios

no vcore boost

qpi pll: 1.2v

ioh lpp: 1.425v

cpu pll: 1.425v

ioh vcore: 1.1v

cpu vtt: 1.2v stable, 1.075 right before bsod

vdimm: 1.63 in e-leet, showed as 1.65 in bios

Thanks in advance for any help guys, I'm brand-new here & hate for my first post to be a "need help" deal but I'm in a bad way after this experience and need the wisdom of more knowledgeable people.

 

[ObscureParadox]

Why exactly did you try to get the chip to run 4.0Ghz @ 1.075v That's overclocking at below stock voltage, that was never, ever going to work, blue screen was always going to happen, i don't know but i would pt money on you've damaged the chip now.

 

[Rastalovich]

x58 mobos are generally quite robust and even if you louse something up pretty badly, they should have a backup bios feature when all else fails.

If you have indeed screwed the cpu, then you're stuck at this point.

However, strip the mobo down, put 1 stick of memory in it, disconnect all the drives - in effect take it down to the least demanding, basic, setup.

Clear the cmos, not by using the button, but find a jumper and check with the manual (find it @ the evga site if you need to) for the clr_cmos pins. Turn the psu off, but leave the mains lead in - stick the jumper on, press the power-on button on the mobo until all lights (if any) go out, and leave the jumper on for.... christ knows how long (10 seconds is usual when you're all in happy-mode). More than 10 mins, hours aren't necessary. If your mobo is designed where it has no jumpers, just the clr cmos button, then you'll have to hold it down for the duration.

After that, switch on the psu, and go for turning the mobo on again. If you get absolutely nothing - you're knackered.

Intel cpus *sometimes* have a manufacturers warranty where they may replace it (if they got any). My qx9650 sensors died and Intel, through Scan, replaced it a long long time after I purchased it.

Be strange for the mobo to be dead on a low voltage thing. If you could get another cpu to test it, great. Doubt it's bust tho. Chances are it's got screwed up microcode thoughts about the 980x.

I'm hoping in your case it's thinking the 980x operates at low volts and needs re-familiarizing.

 

[airdeano]

it the manual http://www.evga.com/support/manuals/files/141-GT-E770.pdf there is power LED, is there power for the board?

there are voltage pads to read voltage to verify current. break out volt/ohm meter and compare.

or you may wanna contact EVGA for a solution.

airdeano

 

[why_sleep]

x58 mobos are generally quite robust and even if you louse something up pretty badly, they should have a backup bios feature when all else fails.

If you have indeed screwed the cpu, then you're stuck at this point.

However, strip the mobo down, put 1 stick of memory in it, disconnect all the drives - in effect take it down to the least demanding, basic, setup.

Clear the cmos, not by using the button, but find a jumper and check with the manual (find it @ the evga site if you need to) for the clr_cmos pins. Turn the psu off, but leave the mains lead in - stick the jumper on, press the power-on button on the mobo until all lights (if any) go out, and leave the jumper on for.... christ knows how long (10 seconds is usual when you're all in happy-mode). More than 10 mins, hours aren't necessary. If your mobo is designed where it has no jumpers, just the clr cmos button, then you'll have to hold it down for the duration.

After that, switch on the psu, and go for turning the mobo on again. If you get absolutely nothing - you're knackered.

Intel cpus *sometimes* have a manufacturers warranty where they may replace it (if they got any). My qx9650 sensors died and Intel, through Scan, replaced it a long long time after I purchased it.

Be strange for the mobo to be dead on a low voltage thing. If you could get another cpu to test it, great. Doubt it's bust tho. Chances are it's got screwed up microcode thoughts about the 980x.

I'm hoping in your case it's thinking the 980x operates at low volts and needs re-familiarizing.

Thanks for the advice Rastalovich, will try this when I get off work = )

 

[why_sleep]

Why exactly did you try to get the chip to run 4.0Ghz @ 1.075v That's overclocking at below stock voltage, that was never, ever going to work, blue screen was always going to happen, i don't know but i would pt money on you've damaged the chip now.

Yes I realize that vtt change was in the lower range, because it was passing all "stability" testing I could throw at it I assumed the chip was just above-average silicon & could handle some tweaking about in those vtt ranges

 

[why_sleep]

it the manual http://www.evga.com/support/manuals/files/141-GT-E770.pdf there is power LED, is there power for the board?

there are voltage pads to read voltage to verify current. break out volt/ohm meter and compare.

or you may wanna contact EVGA for a solution.

airdeano

Yes the board appears to be receiving power, though the LCD post on the board reads a 77 error code, not sure what this stands for.Tried speaking to EVGA but they're blaming the chip off the bat & won't issue an RMA until I try another chip in it-which I unfortunately do not have, nor a volt/ohm meter unfortunately. Perhaps I'll pick one up...thanks!

 

[KING_OF_SAND]

pour milk on it?

 

[why_sleep]

bump

 

[Firat]

Yes the board appears to be receiving power, though the LCD post on the board reads a 77 error code, not sure what this stands for.Tried speaking to EVGA but they're blaming the chip off the bat & won't issue an RMA until I try another chip in it-which I unfortunately do not have, nor a volt/ohm meter unfortunately. Perhaps I'll pick one up...thanks!

According to the manual error code 77=

Detect serial ports - IDE device detection Initialize serial ports.

Are you using any of these ports? Did you try removing the cMos battery for more than 10 seconds?