kelkel1
CPU-Z validation - https://valid.x86.fr/glgb24
 
As you can see, 75°C with less than 1.3VCore.
 
This is with custom loop.
 
Normal or too high?

bob16314
From reading reviews, those processors do seem to run hot.

What CPU cooler are you using?

The only things I can say is try to get the Vcore down or change the Load Line Calibration (Vdroop) so that the Vcore doesn't increase much, or at all, under load.

Get a better CPU cooler and/or apply some liquid metal TIM like Thermal Grizzy Conductonaut between the cooler and processor IHS..And if you really want to, delid the (soldered on) IHS from the CPU die and use liquid metal TIM there too.

Or don't worry about it, the cores won't throttle until 100C, that's the TjMax temp, anything less than that is fine but it's generally accepted that anything more than 10C below TjMax is good.

Nothing you can do about it getting hot when running AVX except to increase the AVX Offset so that it downclocks when running AVX.

Also, just run IBT at a Standard Stress Level and not Maximum. 

And you should be running at maximum cooling, fans and pump at 100%.

Cool GTX

kelkel1
No-one thinks the VSA or VCCIO are too high?
 
I had read somewhere that those voltages can affect CPU temps.

What are your Values ?


VSA Default = 1.05
 
VCCIO Default = 0.95
 
 
 
Good Read: https://xdevs.com/guide/e399ocg/

(excerpt)
 
Usually good rule of thumb is to use around 1.20 … 1.35 V on VCCIO and VSA rails. These voltage levels provide enough margin to maximize memory overclocking for common CoffeeLake CPU. For 24/7 daily use optimal to use 1.25 V or below, given this is enough for desired memory and CPU clocks. If you having stability issues DDR4 overclocking, worth to try various settings, in range from 1.15 to 1.35 V to isolate bottleneck. Some CPUs and memory configurations actually may become unstable when VCCSA voltage is over 1.30 V, so testing both of these voltage might take some time to reach a sweetspot. Memory voltage is also important factor to the system stability. Because this finetuning is dependent on particular CPU sample and memory used we cannot give one set of settings that will work for every case. It’s bit of a process here.
 
 

Cool GTX
yes, I saw the eye test you posted
 
the numbers did not Jump out, though it easier to ask



 
according to xdevs 1.4xxx V  - is on the high side, time to try stability testing at lower range 1.15 - 1.35 V  [Daily use 1.25 V or less]
 
If you do Not know - that article on xdevs was written by the EE on the EVGA Dark Development Team - TiN_EE
 
EVGA Z390 DARK toolbox - OC software, tips and tricks, BIOS profiles, etc

bob16314

kelkel1
No-one thinks the VSA or VCCIO are too high?

I had read somewhere that those voltages can affect CPU temps.

Yes and no.

Your VCCIO and VCCSA is much more than is probably needed, but won't affect CPU temp any (doesn't for me at least)..My VCCIO and VCCSA on [Auto] run 1.424/1.416 respectively..I set them down to 1.20/1.25 respectively and still maintain stability (much less pass POST) for my CPU/RAM overclock of 5.2GHz/4000MHz.

I see your Vcore is hitting 1.4 Max under load in HWiNFO, that's probably more than needed too and why your CPU/Core temps are higher..Try to get that down as much as you can by either changing the Load-line Calibration and/or the Vcore Offset.

bob16314

kelkel1
 
Other than VCore, what else would help with temps?

In your situation, with liquid cooling already, I would grab some liquid metal TIM like in my Post #13 above, the stuff is marvelous..But you need to watch the Vcore under load, that's what matters most as far as voltage goes.