• 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.

10700KF hits 90 degrees with Hyperthreading enabled, on CPUz stress test

Soldato
Joined
30 Jun 2019
Posts
8,110
This is too high, right?

The temps are better with HT off, maxing at 84 degrees Ran it for longer and it eventually hit 90 Celsius.

The cooler is a Hyper 212 Evo.

EDIT - With turbo boost enabled, all core @4.6-4.7ghz.
 
Last edited:
Gonna try reducing the voltage (was upto 1.243v under load) and stress testing it, 1.15v seems to boot.

Update - Cores ranging between 78-82 Celsius so far.
 
Last edited:
This is too high, right?

The temps are better with HT off, maxing at 84 degrees Ran it for longer and it eventually hit 90 Celsius.

The cooler is a Hyper 212 Evo.

EDIT - With turbo boost enabled, all core @4.6-4.7ghz.

Get a better cooler. The 212 is entry level enthusiast at best. I mean its ok but not up to cooling a high end cpu these days imho
 
what are you using your pc for ? many people do these silly tests and in reality literally no one will ever stress their pc like some of these tests . if its for gaming just game check temps. bet its well within normal temps in games.
 
This is too high, right?

The temps are better with HT off, maxing at 84 degrees Ran it for longer and it eventually hit 90 Celsius.

The cooler is a Hyper 212 Evo.

EDIT - With turbo boost enabled, all core @4.6-4.7ghz.

CPU-Z is nothing, if you're hitting 90c in that you're going to need a much better cooler, unless all you do is play games, try encoding a video and watch what your temps do then....
 
Seems fine at 1.15v, it's maxing out at 82 Celsius. I might restrict the AVX ratio if needed.

I may upgrade the cooler if temps exceed 85 degrees under load.

Doesn't thermal throttling generally occur on Intel CPUs at 100 Celsius?
 
Last edited:
CPU-Z is nothing, if you're hitting 90c in that you're going to need a much better cooler, unless all you do is play games, try encoding a video and watch what your temps do then....

I ran it in Prime95 with AVX and it (unsurprisingly) gets really hot, ~90 Celsius.

I had my 'Tau Limit' set to double in the ASRock Bios, which didn't help.

I found that my CPU was using a max of 94 watts while running a stress test in CPUz.

So, I capped my PL1 and PL2 limits to 100 watts in my BIOS. This is basically what happens while running Prime95:

https://i.imgur.com/vb4iPXI.jpg

If I don't do this, Prime95 uses the full TDP of my CPU, upto 125w...

So, a 100w power limit seems sensible for my CPU + cooler. Setting an AVX limit in the BIOS didn't really work properly...
 
Last edited:
Out of curiosity, how much power do the 10900 and the 10900K (at turbo speeds) use under full load in programs like CPUz?
 
what are you using your pc for ? many people do these silly tests and in reality literally no one will ever stress their pc like some of these tests . if its for gaming just game check temps. bet its well within normal temps in games.

To OP, I would say the same, what's it reaching during normal use? Given that normal is different depending on usage.

Also, what are you doing about fan speed control?

The 212 may be old but I think it's still a very well respected air cooler. If you're not pushing max all core overclock then I don't see why there would be any problem, but I'm a bit out of touch to be fair.
 
I'm running it at 4.7Ghz all core, using the turbo boost feature. The fan on the cooler is quite a good one and runs at 2700RPM on load.
 
what are you using the pc for ? that is the question. video rending for eg is going to get it hot gaming so so. for gaming and most tasks the cpu cooler you got is fine.
 
what are you using the pc for ? that is the question. video rending for eg is going to get it hot gaming so so. for gaming and most tasks the cpu cooler you got is fine.

Exactly this... I really dont get why people put their systems through these torture tests if all they ever do is play games. Fair enough if you're encoding videos or some other core intensive work, but even then I would be observing temps using the PC as you would use it, not synthetic benchmarks.

I pretty much only game on mine these days, CB20/23 my CPU will hit 72 degrees after several runs, but in actual real world use it idles at a 34 degree average and never gets any sustained temperatures much above 60 degrees.
 
I'm still running Prime95 tests, lol it's pretty tedious. I'm running the CPU with a 100w power limit to keep temps under control. I've returned my RAM settings to default XMP.

Here's what I've found out:
CPU at auto / stock voltages with turbo boost off is stable
CPU at auto / stock voltages with turbo boost (all cores) on isn't stable

1.2v + turbo boost (all cores) isn't stable
1.3v is stable + turbo boost (all cores) is stable
1.25v is stable + turbo boost (all cores) is stable (5 hours)

Lowest stable voltage:
1.21v + turbo boost (all cores) is stable (5+ hours)

What I don't know:
Does running the RAM at 4133mhz, 1.5v effect stability

It's ridiculous that the CPU isn't stable at auto voltages with turbo (all cores) enabled. I can see the CPU voltage dipping a bit at default / auto settings, from the BIOS.

But when the CPU voltage is set manually, level 1 Load line calibration is enabled and the voltage remains constant in the BIOS.

The weaknesses of Intel's 14nm CPUs and their high power consumption are clearer than ever...

I wonder what the max power consumption of Alder Lake will be, at turbo?
 
Last edited:
What's the point in buying a higher end CPU, and then throttling the crap out of it with power limits and AVX offsets just so it doesn't set on fire? Buy a better cooler and run it as intended...
 
If under normal usage, the power doesn't exceed 100w, I don't see a problem. I'm not going to use an AVX offset, because it doesn't seem to be reliable.

I may buy a better cooler, but only if the voltage pushes the power usage above 100w during normal use and gaming.
 
You've already said it runs cooler with a power limit, so that strongly suggests it was using more power than your present limit before.
 
Oh well, maybe Zen 4 will probably bring down power consumption significantly (5nm EUV can reduce by 30% vs standard 7nm), so maybe we will see 8 core CPUs with max power consumption (Prime95) of 100w-110w. According to this, the overclocked 5800x consumes upto 160w of power in Prime95:
https://tpucdn.com/review/amd-ryzen-7-5800x/images/power-stress.png

I'm doubtful Intel will reduce power consumption to similar levels with Alder Lake, given the approx. 220w max usage of the 10700K in Prime95.

Would be nice to get a stock cooler that can actually handle this kind of power usage (110w).

Update - This is pretty crazy, upto 433w (temps upto 110 celsius) for the 11900K with adaptive boost enabled:
https://tpucdn.com/review/intel-core-i9-11900k/images/power-stress.png

And the Ryzen 5600x consumes just 134w in Prime95!
 
Last edited:
@g67575 As others have said, it's not the most sensible thing to do in running Prime95, on a 10700k, with a 212 and then complain about the temps. Get a decent, up to date cooler like the Freezer ll .

Why are people in this day and age still using Prime95? You're better off using Realbench as that puts a more realistic load on the CPU and also stresses the GPU subsystem at the same time.

Also be careful in making a direct correlation between power usage and heat output. The design of the CPU can also greatly affect this. There is a ~25c difference between my 3950X and my 5800X when running say CBr20 and they both use ~140W.
 
You can run Prime95 with a super duper cooler, doesn't mean it will be stable, even at default / auto CPU voltage settings (as I've found out).

I'll get a better cooler if I need one - with Alder Lake / Zen 4 out next year, buying an expensive new cooler might not be the best decision long term.
 
Back
Top Bottom