Are these temps too high?

[schmohawk]

Hey Everyone!!

Just built a machine with the following specs:

CPU: Intel Core i7-6700K 4.0GHz LGA 1151
MOBO: ASUS Z170-A ATX DDR4
RAM: Kingston HyperX FURY Black 16GB Kit (2x8GB) 2133MHz DDR4 Non-ECC CL14 DIMM
GPU: ASUS GeForce GTX 960 Overclocked 4GB GDDR5
HEATSINK: Cooler Master Hyper 212 EVO with 120mm PWM Fan
PASTE: Arctic Silver 5
SSD: Intel 480GB
HDD: WD Blue 1TB 7200 RPM HDD
DVD: ASUS 24X DVD/R/RW
PSU: EVGA SuperNOVA 650 G1 80+ GOLD, 650W Continuous Power, Fully Modular
CASE: Corsair Carbide Series 200R Compact ATX

Everything is stock configuration, no OCing at all.
I tried the X pattern for thermal paste for the first time, so I wanted to check the CPU temps after installations were done, and this is what I'm getting:

With Prime95 working, blend test:





With Prime95 stopped:



Those temperatures seem very high to me at load.. am I right to be concerned??
Any input is greatly appreciated!!

 

[EarthDog]

Blob in the middie is best... not sure that is the reason but temps seem high.

Update your signature..

 

[schmohawk]

This won't be my PC, so I kept my signature as my machine at home.

To be honest, I've never built a computer without spreading the paste manually before, so after reading this rather exhaustive comparison of every method possible, I decided on the X pattern: https://www.pugetsystems.com/labs/articles/Thermal-Paste-Application-Techniques-170/

Only thing is, I was so concerned about overflow since I'm used to spreading paste all these years, that I kind of re-lifted the heatsink as I was setting it because I couldve sworn some was going to ooze out the sides. Turns out that article was right, and the coverage was almost perfect, but now I have concerns that checking the paste introduced air bubbles.. I get like that.

I'm running into a few forum threads stating that some of these new i7s run a bit warmer and possibly these are normal temperatures?:
example: http://www.tomshardware.com/answers/id-2810472/cpu-temp-range-6700k.html

... then I'm thinking maybe the paste needs it's 200 hours to cure?

... then I start to think maybe I should take everything apart and re-paste the cpu/cooler? I was worried I would end up having to do that.



Thanks again .. any ideas are greatly appreciated.

 

[caddi daddi]

yup, a little high, start by reducing your vcore in the bios.

 

[EarthDog]

Lifting it off and putting it back on couldn't help...correct.

Cure time yields maybe 1C or so...

 

[schmohawk]

Is it typical to need to reduce vcore with this mobo/cpu combination fresh out of the box?

Id hate to start making bios changes if I indeed need to re-do the thermal paste. .. but in the back of my mind my rational thinking is telling me that re-doing the paste is probably overkill and not the source of the problem.

Have you had experiences where core voltage needed to be reduced before any OCing was done? Total default settings, all items purchase brand new.

- - - Updated - - -

Lifting it off and putting it back on couldn't help...correct.

Cure time yields maybe 1C or so...

Are you saying that removing it now and cleaning it off and reapplying won't help much..
..or that having done it during the build was a bad move and it should be re-done now.

Apologies for not following, not much sleep today.

 

[caddi daddi]

prime 95 after 26.6 you should not use with intel.
try it with inel xtu. with later versions of prime they produce huge amounts of heat.
and do reseat the heatsink, you might have trapped some air.

 

[schmohawk]

was afraid of that.. i get antsy when handling the cpu.
I'll make those software changes after reseating the heatsink and post results.

Never had to re-do a heatsink before.. is it advisable to leave the mobo in the case (disconnected power of course)?

Also, some guides state to completely remove the cpu from the mobo to clean it off with q-tips and rubbing alcohol.. i was hoping to get away with cleaning it without having to disconnect it. Any experiences/advice regarding cleaning and reapplying paste/heatsink?

my hands get a little shaky when handling the cpu out of the mobo.

 

[caddi daddi]

just remove the heatsink, leave the poor little cpu in the socket as much as possible.
to clean mine I just use coffee filters and alcohol.

 

[Lochekey]

I do my cpu while it is still in the mobo unless there is an excessive amount of Tim that squeezed out and is under the hold down bracket. If that is the case then I pull the cpu and give it a good cleaning, especially on a system with artic silver as it is conductive.

- - - Updated - - -

Ninja caddi strikes again

 

[schmohawk]

i read that with a sigh of relief.. i usually hold my breath when connecting the cpu

- - - Updated - - -

ill check for seepage

 

[schmohawk]

ill check for seepage

Houston we have a problem..



That X turned into quite a thick schilacking of thermal paste..



First the dry clean with unbleached coffee filters-



More discerning light highlighted some residue-



99% iso cleaned the heatsink.. took many passes to get the coffee filter to come up clean. Longer than I thought-



Carefully 99% iso cleaned the cpu. Wasn't as stubborn to get clean as the heatsink-



Now I'll carefully lift those brackets and get any residue underneith. I'll probably leave the cpu seated in the motherboard though as I don't yet see any areas needing cleaning that would require me to lift the chip.




Taking that heatsink off was quite the moment. It really didnt want to let go and I knew too much pressure would force the cpu to try and pop out. I had to keep gently twisting which I think caused the majority of the overflow on the sides. Still, even if none of it overflowed, that looks a bit thick for thermal paste.

What say ye? Give up on the X and go with the pea this time?

 

[schmohawk]

...go with the pea this time?

.. the Metallic Pea perhaps?

 

[schmohawk]

Cleaned underneith the bracket-




Locked it in and gave it one last pass with 99% iso.




I feel a lot better about that now!

..Now if only I felt as good about just a small rice grain sized spot of thermal paste. I guess old habits die hard and my subconscious is dying to spread the paste. Even though I totally believe in these results- (from here: https://www.pugetsystems.com/labs/articles/Thermal-Paste-Application-Techniques-170/ )



I'm also thinking the risk of to much paste when drawing the X is outweighed by the slightly better temps it had over the rice grain / pea size dot.
But man.. that is not a lot of paste at all! -

from this: (their example)



to this:




I know the thermal paste 'dot vs whatever' has been beaten to death..

I guess I'm just a bit more concerned with this PC since it'll be used by my father for his business.. and I may not be on hand to immediately assist if he has any issues in the future. Just want to maximize the longevity of this system by as much as possible.

Any words of confidence to convince me to forget the X and just do the dot already?

 

[schmohawk]

Now I'm more conflicted than ever..

I went with the majority opinion and tried my best at a grain of rice / pea size dot of thermal paste-



And my temperatures got worse!

Even though I know Prime95 runs hot on this processor, I wanted to peek and see if there was a difference in the temps it displayed last time.. and there sure was a difference, now they're higher!



I had a feeling I should have left well enough alone. That X really was a better application method.

caddi daddi recommended Intel's Extreme Tuning Utility, so I installed and ran the 5 min stress test with these results-



Then ran it with HWMonitor-




What do you think?

TBH im pretty bummed that the temperatures went up.. I'm not really looking forward to having to do that again for the third time, and Im getting that feeling like I should've trusted myself.

 

[caddi daddi]

what?? sometimes I have to remount a cooler 3 or four times to be happy.
in your pic of the chip with the tim spread, the areas you marked are over air, the cores are a slug down the center of the heatsink from the hinge of the socket to the latching screw.

 

[EarthDog]

 

[schmohawk]

what?? sometimes I have to remount a cooler 3 or four times to be happy.

That's what I needed to hear!

Also, those last examples with the circles weren't mine, they were images from the study I linked.

So here's what that pea sized dot turned into:





And while I do agree that the cores generating the heat are only in a small area in the center of the metal enclosure.. that entire plate of metal is absorbing that heat, and I'm not 100% convinced that thin paste contact with the entire surface makes no difference.

Either way, from my own personal experience with both the dot and the X, I went again with the X this time, although using less paste-



Ran Intel's stress test and these are my results:



Keeping in mind that I don't plan to oc this cpu, but also that it'll probably be a 24/7 office machine.. Do these temps seem adequate?

 

[adantier]

try using this paste - Shin-Etsu X23-7762. it is the best paste i've tried, and it could be paste paste your using is the problem (bad company, bad batch, etc)

 

[JrClocker]

The "pea size" dot you tried was not in the enter of the CPU...that's why the blob didn't squish out to cover everything. See ED's comment about having to do a mount a few times until satisfied.

You also need to tighten down the heatsink in a diagonal criss-cross pattern. If you tighten clockwise or counter-clockwise, the paste will not spread out evenly.

Also, the purpose of the paste is to fill small voids in the metal-to-metal interface. Not to provide a full interface between the two pieces of metal...in other words, more is not always better.