Help diagnosing BSOD 0x00000124..?

Try in small increments,try with 1.55v dram first as it could be that,it won't hurt your ram,I had some vengeance and it needed more than 1.5v to be stable

Thanks :)

Would that cause BSOD under prime?

I suppose prime does stress memory as well doesn't it.
 
The memory is now on 1.545v (I had the option of 1.545 or 1.560)

Currently stressing on prime95 large FFT while I go make coffee...

*edit* i'm back, it hasn't blown up yet which is nice :rolleyes:

Switched to blend because it 'tests some of everything'.
 
Last edited:
I got a BSOD after an hour of blend (while watching YouTube in 1080p at the same time), I have put the DRAM voltage back to 1.5v and upped the VCORE to 1.368v. As it turns out my VCORE was on 1.360v not 1.344v and CPU-Z was just reading it incorrectly.

Stressing with prime again.

My VCORE is set to offset +0.60v (1.368v) is that right? I remember it originally being on fixed, but someone told me to set it to offset.

*edit* When my VCORE was set to offset +0.55 (1.36v) in the BIOS, CPU-Z read it as jumping between 1.337v and 1.344v. Now it is set to offset +0.60 (1.368v not sure how that computes, but it is what it says) CPU-Z reads it as jumping between 1.344v and 1.352v... CPU-Z is so helpful :rolleyes:

My temperatures have risen by about 2°C now, I suppose that is normal after a voltage increase.

*edit again* I BSOD'd after a mere 15 minutes that time. So now I have tried reducing the VCORE rather than increasing it. The VCORE is now set to offset +0.50v (1.352v).

Stressing with prime again.

*edit again* I BSOD'd after just 6 minutes that time.

I will take some BIOS screenshots to show you guys what settings I have, they will be up later.

Back on the original 1.36v now.

So after adding 0.045v to the RAM the problem persisted, after adding 0.008v to the VCORE the problem persisted, and after reducing the VCORE by 0.008v the problem persisted. Might I need to do both at the same time? I somehow doubt that a 4.7GHz overclock with 1.360v would be stable for a year, and then suddenly 4.7GHz becomes unstable with even 1.368v.
 
Last edited:
does it only bsod in games?

how much vccio or vccsa voltage is being used? are you enabling x.m.p for the memory? see if that helps as it sets up the dram/vccsa voltages for you
 
does it only bsod in games?

how much vccio or vccsa voltage is being used? are you enabling x.m.p for the memory? see if that helps as it sets up the dram/vccsa voltages for you

No, it BSOD'd a few times on Prime95 as I mentioned before, and it also BSOD's in games and while playing youtube video's sometimes.

I don't know how much vccio or vccsa voltage is being used, I don't even know what they are if i'm honest, the only things I understand are FSB, multiplier, VCORE voltage and DRAM voltage. Everything else it a mystery and I just do what i'm told regarding the other values.

Surely the length of time between starting prime95 to BSOD changing relative to my VCORE suggests that it is the CPU overclock though..?

I will sort out the X.M.P profile and report back after some prime95 blending.

Is there anything that I need to do other than simply 'enable' it?
 
Last edited:
enable it and it will list the profiles,each profile will set the speed of the memory,some have more than one profile some just one

its either ram or cpu voltage,one or the other or both

video or gaming/crashing/lockups ect usually points to ram imo,but could be wrong

blend test in prime tests a lot of ram aswell

vccio is for the integrated memory controller inside the cpu,vccsa is only needed if raising the bclk which you dont need to do on ib

can be dram too low aswell 1.55v tends to be the sweetspot
 
enable it and it will list the profiles,each profile will set the speed of the memory,some have more than one profile some just one

its either ram or cpu voltage,one or the other or both

video or gaming/crashing/lockups ect usually points to ram imo,but could be wrong

blend test in prime tests a lot of ram aswell

vccio is for the integrated memory controller inside the cpu,vccsa is only needed if raising the bclk which you dont need to do on ib

can be dram too low aswell 1.55v tends to be the sweetspot

OK thanks a-lot, you have been a great help so far :D

Hopefuly we can get this sorted out soon :)

My chip is a Sandy not an Ivy though, but I guess its the same regarding not having to change the base clock.
 
Just went into the BIOS. The XMP profile was set to 'auto' (DDR3 1600, 9, 9, 9, 24, 1.50v) and every setting thereafter was also set to 'auto'. There was another profile titled 'profile 1' which I could select as the XMP profile instead of auto, but when I selected it, nothing changed, all settings remained the same.
 
your ok to use upto 1.575v dram voltage on sb chips,so id try atleast 1.55v and then test and raise cpu v if needed later,maybe one/two clicks more than stock 1.05v vccio

but usually thats fine on stock 1.05v
 
Hi, you say 8GB Vengence Ram, your safe upto 1.6v try increasing ram voltage.


Corsair Vengeance 8GB (2x4GB) DDR3 PC3-12800C8 1600MHz Dual Channel Kit (CMZ8GX3M2A1600C8)
1600MHz RAM Speed, CAS 8-8-8-24 Timings, 1.5-1.6v VDIMM, Lifetime Warranty with Corsair.

Corsair Vengeance 8GB (2x4GB) DDR3 PC3-12800C9 1600MHz Dual Channel Kit (CMZ8GX3M2A1600C9)
1600MHz RAM Speed, CAS 9-9-9-24 Timings, 1.5-1.6v VDIMM, Lifetime Warranty with Corsair.
 
How does this look? Anything obviously wrong?

IMAG0405_zpsb6c70b2e.jpg


Apart from the DRAM and VCORE voltages which might need adjusting.
 
Last edited:
could try with vtt at 1.06-1.1v

and enable power saving,it shouldnt affect stability

you can also try one level higher load line calibration and less offset say +0.040v or +0.045v you would need to test
 
i had a similar issue while trying to overclock using offset. turns out my board and cpu just doesn't like being overclocked by using the offset method so i switched to manual overvolt and it's been sweet since!

Thats ivybridge 3570K at 4.7ghz with a vcore of 1.395 and high LLC which keeps my voltages up at around 1.4v underload.
 
could try with vtt at 1.06-1.1v

and enable power saving,it shouldnt affect stability

you can also try one level higher load line calibration and less offset say +0.040v or +0.045v you would need to test

If enabling power saving would not affect stability, what is to be gained from doing this..?

Also what is 'load line calibration'? and how will it affect the overclock/performance etc?
 
it might do at 4.7ghz idk but only way to find out is to try,no issues for me at 4.5ghz and 4.6ghz with offset and power saving enabled

it just downclocks the cpu speed and voltages at idle and switches to full speed/voltages at load less than 1v at idle

loadline calibration adds a touch of extra cpu v at load to compensate for cpu voltage droop at load,say you set 1.25v cpu at load it would be 1.23v ect thats the vdroop

each level of llc adds more voltage so you can use higher level of llc and less offset or vice versa
 
I just changed it to 1.36v fixed rather than offset, and the system bluescreened before I could even log in. Needless to say i went back to offset again.

I am now running blendf on prime with 1.064v VTT and 1.56v on the DRAM.

Lets see how it goes.

Is there any reason why this has made my temperatures rise considerably? I was looking at 77 accross the board before, now im looking at values ranging from 77 to 84.
 
Last edited:
After running blend with the VTT at 1.064 and the DRAM at 1.56, it did bluescreen, but it took a full 9 hours to do so, while playing HD video at the same time. I have now put the VTT up to 1.077 in the hope that it will help.

Any idea why my RAM etc. might suddenly be more power hungry? (I say suddenly, it's more something that happened gradually I suppose)
 
Last edited:
it could be down to you running 4.7ghz and putting more strain on the ram? idk but 1.56v dram voltage isnt a lot so i wouldnt worry to mch,intel safe limit is 1.5750v and ive seen some use more,it just depends on your ram and your cpu's memory controller
 
it could be down to you running 4.7ghz and putting more strain on the ram? idk but 1.56v dram voltage isnt a lot so i wouldnt worry to mch,intel safe limit is 1.5750v and ive seen some use more,it just depends on your ram and your cpu's memory controller

There are people all over these forums (and of course elsewhere) running respectable (or in some cases completely insane) overclocks alongside 1600MHz RAM though :(

Particularly on Ivy and Sandy i5K's and i7K's.

Anyway, I got a BSOD with 1.077v VTT and 1.56v DRAM after an hour, so i'm back to 1.064v VTT now.

I have a suspicion that it might be due to something at the top of the BIOS screen which I posted before:

IMAG0406_zpsccb9b47a.jpg


The power limits perhaps..?

I'm not very adept ad this, so I might be talking rubbish of course :rolleyes:
 
im not sure on that board,you could try raising power limits a little see if that helps,you can still use a touch more dram v upto 1.575,

disabling cpu spread spectrum might help also
 
Back
Top Bottom