hight temps for 8700k

Oct 23, 2018
1
0
10
Specs i7 8700k + Prime z370-A motherboard

Hey need a little help after i installed my new h150i pro aio i tryed to overclock to 4,7ghz is it normal for it to be idle 38c (sometimes spikes to 50c) and under 100% on aida64 its on 92c+?

or did i got a broken aio ? btw it makes strage noise

so the only tweaks i did in bios i manualy added 47 clock and left the rest on auto
 

CompuTronix

Intel Master
Moderator
akzumon,

On behalf of Tom's Moderator Team, welcome aboard!

Is "idle" really idle? Only 1 or 2% CPU Utilization in Windows Task Manager?

There are many variables, so please be more specific.

What is your ambient temperature? 10°C? 22°C (normal)? 30°C? 40°C? Always provide ambient ... it can be a huge variable, so it's critical information that everyone should always include.

92°C is too hot. It's not sufficient to simply say that you ran AIDA64. What exactly did you run?

AIDA64 has 15 possible test selections involving the processor, which yield 15 different Core temperatures. Only one of those selections is actually a valid 100% TDP workload.

(1) CPU - < 70% Workload
(2) FPU - > 110% Workload
(3) Cache
(4) RAM
(5) CPU & FPU ~ 100% TDP Workload
(6) CPU & Cache
(7) CPU & RAM
(8) FPU & Cache
(9) FPU & RAM
(10) Cache & RAM
(11) CPU & FPU & Cache
(12) CPU & FPU & RAM
(13) CPU & Cache & RAM
(14) FPU & Cache & RAM
(15) CPU & FPU & Cache & RAM

We have "Stickies" at the top our Forums, which are readily available references provided for everyone's benefit. Give this one a read; it will show you how to correctly perform valid load and idle thermal tests: Intel Temperature Guide - http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

Once again, welcome aboard!

CT :sol:
 
92C would seem high even for MCE-enabled all core at 4.7 GHz, IMO....

As for CPU loading, I've found in my experimenting that even Intel's Extreme Tuning Utility stress testing (CPU or RAM stress tests) generates CPU loads that caused temps pretty much identical to those of Prime95/small FFT , in my case peaking at about 74C on my 7700K with MCE enabled... (the cpu stress stress test is more consistent with the core speeds/temps, but, both will hit 74C , at least on my system. (I'd naturally expect an 8700K to run a tad warmer, perhaps 80-83C?)

The 'hearing weird noises' from pump is concerning, however.

But, certainly you could try double checking that the pump is evenly locked/secured down onto the heat spreader, not *too much* or too little paste, etc..and making sure pump is working at full speed /w medium high fan speeds, etc. , always)

With MCE enabled, WIndows power plan set to Balanced, you can leave multiplier on Auto, sync all cores, and if temps allow, MCE mode alone will run all cores at 4.7 Ghz if it can anyway during heavy cpu load, no need to set multiplier manually... Check to see what core voltage is being applied during all-core 4.7 GHz operation
 

CompuTronix

Intel Master
Moderator

Which version?

As I know you're well aware, temperature differences between "later" AVX versions, and the non-AVX version 26.6 can be huge. Many of our members and readers are very confused about running "Prime95" because they read so many conflicting threads about it, where the version dependent nature of the AVX issue is never clarified.

This is why I always make it a point to specifically state which version I'm referring to when discussing Prime95. All too often, whenever users see "Prime95" in a thread, many shy away from it because of the bad press it gets due to so many misinformed users repeating what they've read from other misinformed users. If we don't differentiate between versions by clarifying, then the confusion just continues to be perpetuated.

So once again, for everyone's benefit, and FOR THE RECORD, THERE IS NOTHING WRONG WITH PRIME95 VERSION 26.6 SMALL FFT's. It's actually the most ideal thermal test available, because it's a steady-state 100% TDP workload. There is no other test that so closely replicates Intel's proprietary workload, which is used for validating thermal specifications as described in the datasheets. P95 v26.6 Small FFT's is not an overload OR underload; it's the proper workload for thermal testing.

Prime95 is just one of many points of confusion which occurs on our forums. We do our best to ensure that our content is as clear and accurate as possible. Fortunately, we have a few members who help us to accomplish this task by correcting misconceptions, which goes a long way toward dispelling ongoing fallacies. mdd1963, as a prolific member who offers good advice, I respectfully ask when mentioning or discussing Prime95, that you please help us to un-confuse our less informed members by clarifying thread content. Your help would be greatly appreciated.

CT :sol:
 

TRENDING THREADS