• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

At my wits end - 3930k help

Take the case side off and stick a fan blowing sideways onto the CPU heatsink see if that helps? it might be an airflow problem with Crossfired GPU's and too much warm air is building up inside the case.

I'm at [email protected] with Noctua NH-D14 2011 and Silverstone FT02 and get 70-75C under Prime95, but this case has amazing cooling tbh and my GPU is the exhaust type whereas yours aren't?
 
eyetrip
Don't beat yourself up too much these chips generate a lot of heat.
I use a D14 on mine to keep the temps it check,
If you think it is hot now wait till you OC it.

I also have a 3820 and that runs pretty cool, about the same as a 2600 but the hex core is a furnace.
 
Core temp is reporting same temps as Aida64 - i.e. more realistic.

Will be sticking with those 2 for readings.

Currently on 1.28V at 4.5GHz. Temps under 80, but need to bring down more.
 
Must be something to do with the mount, i know you have tried it several times but 70 degrees idle is insane. Are all the posts fully screwed in for the cooler, not using too much tim or applying uneven pressure on the ihs? From my experience these chips run pretty cool, i get lower temps at 4.7ghz than i did with my i7 920 d0 @ 4.2ghz and the 3930k has 2 extra cores and more cache.
 
Must be something to do with the mount, i know you have tried it several times but 70 degrees idle is insane. Are all the posts fully screwed in for the cooler, not using too much tim or applying uneven pressure on the ihs? From my experience these chips run pretty cool, i get lower temps at 4.7ghz than i did with my i7 920 d0 @ 4.2ghz and the 3930k has 2 extra cores and more cache.

Wasn't idle temps that were over 70 but load temps, which I have now figured is some programs basically reporting temps 10c or so higher than they are.

That's not that bad, so you were using CPUID HM before and now with AIDA/CT it's lower?

I've always used Coretemp.
Yeah, correct - CPUID HW monitor and HWInfo64 are the two I was using intially.

Wish I had tried other programs earlier - wasted hours hours today mucking about with coolers :(

Though, to be honest, don't think my last seating was actually a great job, as I was just so fed up by that point. May do one last re-seating of the Flex tomorrow, then that's it.

Hoping to get a good nights sleep and hopefully spend tomorrow actually playing some games...

PS
Hmm, didn't actually post this earlier when typed it - so much for good nights sleep :(
 
Hello
After a fresh install of my fx8120 on a new motherboard I also had problems with heat (20c hotter than on old motherboard) but after a week or so my temperatures came down back to normal and I was using mx-4 compound so it might be just that it's settling (??)
 
Currently at 1.25V under load at 4.5GHz ( Is that good?), and been stable for many various runs of IBT (max. high, very high).

One core has peaked at 77C, other cores range from 64 to 74.

But I still need to re-check seating of cooler. Probably get that out way when finished my coffee.
 
I get quite a large spread between the temps of the individual cores on my 3930k,
Dont reseat the heatsink too many times to cure it as i believe they all have this trait.

Nice low vcore there, well compared to my C1 chip.
 
Yeah, I wasn't worried about reducing the variation in temps. Just wanted to make sure it was seated decently, as by the time I did the last seating, I was just fed up of mucking about with coolers.

Looks like re-seating has only helped by a degree or two, but I will take what I can get :D

Peak have seen now is 76. If I can keep the 4.5GHz at 1.25V, then I will be happy with the current temps.

Although, I suppose I could try an even lower vcore....... off to do that now :D
 
When said was 1.25V it actually alternates between 1.248V and 1.256V

Just dropped another 0.01V and is now fluctuating between 1.240V and 1.248V

Currently running IBT.

To be honest, I would have expected it to BSOD by now. Seems like very low voltage for 4.5GHz.

IBT speed (Gflops) is just in excess of 150 - is that about right?
 
Dropped another 0.01V (so now between 1.232V and 1.240V), and running another "very high" IBT. Guess I will jsut keep dropping it til I get an error on "very high", then up it back the the last voltage and then do a max test, and hopefully passes :)

Edit: Damn, got an error. So adding another 0.01V and doing max test.
 
Looks like 1.248/1.256V is where it's at for this CPU at 4.5GHz.

Still very happy with that as I was expecting to have to go north of 1.3V, which would have been unlikely given temps.

Looks like for once in my life I got a really nice chip. Perhaps one day I will get better cooling and push for 5. I imagine this cooler would probably cope with Ivy-E though. Will see how long it is before I get bored :D
 
Some useful information here thanks :) , will be tackling my 3930K later today :)

Good luck, and let us know how you get on :)

Still got one voltage to play with, and that is VCCSA, which is on auto and the board sets quite high.

Hopefully can drop that quite a bit and maintain stability. Not sure it will do anything for temps though.
 
When said was 1.25V it actually alternates between 1.248V and 1.256V

Just dropped another 0.01V and is now fluctuating between 1.240V and 1.248V

Currently running IBT.

To be honest, I would have expected it to BSOD by now. Seems like very low voltage for 4.5GHz.

IBT speed (Gflops) is just in excess of 150 - is that about right?

With HT on and running at 4.5Ghz 150 is about right.
It can vary dependent on memory speed.
at 1333Mhz about 145
at 1600Mhz about 150
at 1666Mhz about 151
at 1866Mhz about 153
at 2133Mhz about 154
setting 2n on the memory will loose a little but might be more stable.

BTW 100x45 is a little slower than 36x125 but the difference is very small.
Probably because 36 is more of a syncronous ratio than 45
 
Just still testing lower volts on my ram and it seems that dropping from 1.465V to 1.425V has dropped CPU temps a couple of degrees... didn't think it would affect that.
 
With HT on and running at 4.5Ghz 150 is about right.
It can vary dependent on memory speed.
at 1333Mhz about 145
at 1600Mhz about 150
at 1666Mhz about 151
at 1866Mhz about 153
at 2133Mhz about 154
setting 2n on the memory will loose a little but might be more stable.

BTW 100x45 is a little slower than 36x125 but the difference is very small.
Probably because 36 is more of a syncronous ratio than 45

Think am going to avoid using bclk straps. Used them on 3820, and it disabled the voltage drop at idle. Not sure it will do same on 3930K, but taken a bit to get this OC where it is, so think will stick with it :D

And good to know am on track with gflops throughput :)
 
Yes you don't have any choice with the 3820 if you want clocks >4.4Ghz
It's good that 2011 lets you have the option of BCLK tuning even if you don't always require it.
 
Back
Top Bottom