3570k at stock using Noctua DH-14, do temps look ok?

deankenny

Distinguished
Apr 29, 2013
102
0
18,680
So ive been trying my little heart out to overclock at 4.6 with my new beast of an air cooler, but the volts required to keep stable are slightly above 1.3v vcore to see any stability, of which 10 min prime test pushes temps into the high 80s :( a little dissapointed, and after 4 re applications of thermal in different methods, i decided to put everything at stock and check my temps after 10 mins of prime blend test to see what im working with.

Here is a screenshot of the temps with the CPU all stock, is this normal, above average or below average at stock for this cooler?
832767a980dcbf20a31e5b97751a1ff6.png
 
this is why IB doesn't overclock as well as SB. it's the stupid thermal grease inside the heat spreading "lid" on the chip.

You can *delid your chip for a massive decrease in temps or just leave it be. either way those temps are fine for an intel. You're not in danger with intel unless you're pushing 90C-95C

*warning: delidding a chip does invalidate any warranty you might have on it
 

deankenny

Distinguished
Apr 29, 2013
102
0
18,680
i understand that its well within the limits by a long margin, and so it should be with a £65 cooler on it at stock settings.

My question was with such a good air cooler are these the kind of prime max temps a performance cooler should be getting, or should this cooler with this chip at stock be getting better lower temps?
 

ittimjones

Distinguished
Oct 1, 2012
1,003
0
19,460
It's a great cooler. Mine works great and I never hit 80C, even after hours of Prime95.

It's either that crappy thermal grease under the heat spreader that Intel used, or u could re-apply ur's and re-seat the Noctua, but since you've done that so many times, it must be Intel's fault. sorry to hear. maybe sell it and get a Haswell when u can.
 


like all cpu coolers your ambient room temp and how much solid ventilation your case has will both largely determine if you're getting good performance or not.

it does look hot, but i somehow doubt it's your cpu cooler. it's likely the thermal grease inside the lid, or the other two factors i meantioned.

here is a test you can do, open the side of your case and stick a room fan into it. see what happens to your temps. if your temps drop by more then 5C you have a case airflow issue.

Check what your ambient room temps are as well

also, understand that heat pipes work better in certain orientations. you might get an extra 5C by turning your cpu cooler 90 degrees (i think the Noctua is supposed to be oriented to blow out the back of the case, not the top. so if it's facing "up" turn it 90 degrees to the left.

so there are 3 suggestions to get cooler temps. 1) case airflow. 2) orientation of the cpu cooler. 3) delid the cpu

in the end though what separates a great cpu cooler from a mediocre one is under load. the higher the strain the better it will get rid of heat. Most cpu coolers push out the same temps at rest. its under load the great ones shine.
 

JeffJB88

Honorable
Apr 27, 2013
2
0
10,510
I have the same cpu and fan. Running stock on my asrock extreme4 z77 IBT Very high will produce temperatures in the 66C range at stock settings. Keep in mind at stock voltage it is around 1.2V~ (cant remember off the top of my head). Also Asrock z77 also has the problem of reporting lower voltage than what is supplied. Some users measured up to a 0.1V higher voltage at the output capacitors than what the bios or other utilities are reading.

With that being said I have played around with case air flow several times untill i found an optimum fan orientation in my NZXT phantom mid tower case. The way I know it is optimum is that idle now will drop down to 25C with ambient at 24C (around 76-78 F) and load is lower as well although I can't remember the exact number.

I currently have mine running with mostly stock settings but PLL 1.586V (lowest it can go in bios), CPU offset at -0.200V, LLC Auto (level 5), vram 1.4V (gskill sniper low voltage series supposed to run at 1.25 but is not stable that low). Load temps under IBT very high don't exceed 55C with fan settings on low. Load voltage is reported at being just under 1v. (around 0.98V).

I have spent a lot of time finding stable overclocks at 4ghz but not too much on max overclocks. I do know that for my cpu, it will continue giving WHEA errors even at 1.3-1.32V at 4.5GHZ. At that point I gave up for now as temperatures were reaching 90C under IBT very high and I didn't feel like bumping voltage any more.
 

deankenny

Distinguished
Apr 29, 2013
102
0
18,680
Thanks Jeff, even a similar case to mine, nice comparison than you.

Im at 4.5 quite comfortably, prime been running for 20 mins or so, and no whea errors and my temps are below, but 4.6 requires a huge voltage increase to get stable.
11ee67ed2726fce4c24211b552839d6d.png
 


generally when you start to need to push "exponentially" power through a chip to balance an OC, is when you've reached the end of your overclocking.

i usually step it back a fraction to lower the temps and power draw and use that as me "day to day"

 

ihog

Distinguished


It's not the thermal paste. It's the glue keeping the IHS on the PCB.
 

ihog

Distinguished
Cutting the PCB with the razor, scratching the die (with anything), not setting the IHS centered enough when seating the processor (since the IHS won't be glued to the PCB anymore).

And yeah, if it dies for any reason, you're screwed.

I delidded mine, and I ran into no problems. Ya gotta make sure you clean off as much of the glue as possible.