Kernal-Power Problem

[ncncore]

So my pc started messing about a bit the other day while gaming giving me a BSOD with the error memory management. Now it doesn't go to BSOD after updating my Bios/OC profile. However randomly it now just locks up and all I can do is restart to sort it out. I have noticed that when it locks up completely I get a Kernal-Power (43) Critical warning. Any ideas on what might be the cause?

Here are my specs -

i5 3570k
2x GTX 680's
Corsair AX 750
16Gb 1866Mhz Corsair vengeance
Windows 8.1
1 ssd
2 normal hdds usually on complete power save

OC details -

pci/cpu clock set to 100
cpu multiplier set to 42x
vcore voltage response set to fast
pwm phase control set to extreme performance
cpu voltage to normal
dvid to -0.015v
loadline calibration to medium
all power saving set to on
so now I get a vcore of 1.284v under stress (Prime 95 in extreme)

Finally here is a screen grab from the latest error, this happened while just browsing forums on the net so not even pushing anything -

 

[Josh]

I had the same thing, I stepped all my overclocks back from 4.2 - 3.7 and not had an issue since.

Think for me it was an unstable overclock and a PSU that couldn't cope.

 

[Havana_UK]

I've had the same problem previously which I put down to an unstable overclock (which eventually turned out to be most likely the motherboard's power stage). A kernel power event isn't actually the problem itself. That simply means that something's gone wrong causing an unexpected power-down (in this case, you forcing a reboot). You need to find the other events (critical, error or warning) that happened immediately prior to the shutdown. You'll actually find that the time stamp on the kernel power event is when the PC restarts Windows.

 

[ncncore]

Got you! It's just annoying as the PC has been fine for over a year and now it's started to mess up.

I think I will see if there is a bios update for the mobo and reset the oc see if that sorts out the issue.

I will check the event viewer and see if there was anything else.

 

[ncncore]

Ok so I have switched my BIOS over to the stock one and everything seems completely fine now, cards running nicely, cpu sitting nicely at a 3.8Ghz and no lock ups.

So do you think it's the OC that went bad or isn';t set up correctly or the PSU can't cope, even though it has been fine for over a year?

 

[Havana_UK]

Ok so I have switched my BIOS over to the stock one and everything seems completely fine now, cards running nicely, cpu sitting nicely at a 3.8Ghz and no lock ups.

So do you think it's the OC that went bad or isn';t set up correctly or the PSU can't cope, even though it has been fine for over a year?

OC's don't really "go bad", but what was once a stable OC may become unstable for a variety of reasons such as the CPU temperatures being higher than they used to be (due to higher ambient temperatures or dust build up, for example), the CPU becoming overstressed or power supply issues (and this could be the PSU or the motherboard's power stage).

You haven't given any details of your motherboard. What make/model is it?

Your PSU should be fine running that setup. GTX680s (in stock guise) are roughly 225W I believe, and the CPU, even with a modest OC, shouldn't exceed 100W, so that's 650W. I'm not familiar with that PSU, but I should imagine 100W of overhead (if the 12V rail can deliver 750W) should be sufficient for your other components.

If it is a power issue causing the instability it could be the PSU because of ageing capacitors, or it could be the motherboard's power stage has weakened or was never really up to the job in the first place.

 

[ncncore]

Oh right, I see what you're saying.

For info the mobo is a gigabyte z77x-up5 th. So not a cheap one really.

Having played a few games since putting it back to stock I have noticed no real difference in performance so may just leave, it boosts to just over 3.8Ghz anyway so 4.2 isn't that much over.