cancel
Showing results for 
Search instead for 
Did you mean: 

CPU frequency/ voltage drops no matter what I do.

TheOnlyDoor
Level 7
So here are my specs for starters:
ASUS z390 Maximus XI Hero ( WiFI)
i7 8086
EVGA 2080Ti XC Ultra
32 GB Corsair vengence DDR4 @ 3200
Corsair H115i pro RGB
Samsung 960 pro M.2 2TB
Corsair AX 1500i PSU
NZXT Phantom ATX full tower

Just built this a few days ago, started with this OC tutorial https://www.youtube.com/watch?v=jOXH2YIMi2Y&t=2527s
I followed this tutorial to the letter in UEFI. The host never turns off speed step or any of the C values. when he tests his OC CPU Z always maintains the voltage and frequency during load.
I set mine initally to 5 ghz, 1.3core v (manual). After boot and at idle , CPUz shows 5000mhx and 1.3v. However once I load P95 or Realbench within minutes the vcore drops from 1.3 to 1.28, then to 1.25...etc. followed by frequencies dropping from 5000mhz to 4.8 then 4.4 bottoming out around 3.8. (30 minute Realbench). Temps never get beyond 80-85c even for the minute or two the OC remains at set voltage and frequency.
I've tried adaptive voltage, manual voltage, turning off speed step and all the other ones in that list of 5 or 6. I've tried the predefined OC profiles in Bios........nothing will give me a steady Voltage/Frequency during stress tests.
Dont know what else to do.
5,701 Views
7 REPLIES 7

Arne_Saknussemm
Level 40
Kind of sounds like heat is throttling the OC somewhere...

Is your cooling working as it should? is one of the pipes from the cooler warm...VRM temps OK?

Mr__Fox
Level 12
Have you maxed out the power limits? What do you have set for LLC?

Run ThrottleStop and see if the "Reasons" light up red or yellow. Or, use XTU benchmark and see if the Power Limit or Current Limit bar turns yellow in the monitoring area.
Wraith // EVGA Z690 Dark K|NGP|N | 13900K | RTX 4090 Suprim | 32GB DDR5 Hynix A-DIE @8200 | SuperNOVA 1600 P2 | HC-500A Chiller | MO-RA3 360 || DG-86
Banshee // ROG Z690 Apex | 13900KF | RTX 3090 K|INGP|N | 32GB DDR5 Hynix A-DIE @ 6800 | Corsair RM122x SHIFT | XT-45 NOVA || 5000D Airflow
Half-Breed // Precision 17 7720 | 7920HQ (BGA filth) | Quadro P5000 16GB (MXM) | 32GB DDR4 || Grade A Off-Lease Refurb

HiVizMan
Level 40
Throttling is happening by the sound of it. What airflow do you have over your VRMs? Place a directional fan at the VRMs and see if that makes a difference. With AIO water cooling there is very little airflow over the other components and heat is not your friend. Stuff is designed to protect your system from real harm. Which is what I am suspecting happened here. But I would need to see your complete BIOS to have a more clear picture.

What changes did you make to your bios?
To help us help you - please provide as much information about your system and the problem as possible.

HiVizMan wrote:
Throttling is happening by the sound of it. What airflow do you have over your VRMs? Place a directional fan at the VRMs and see if that makes a difference. With AIO water cooling there is very little airflow over the other components and heat is not your friend. Stuff is designed to protect your system from real harm. Which is what I am suspecting happened here. But I would need to see your complete BIOS to have a more clear picture.

What changes did you make to your bios?

So.....um......turns out one of the Rad fans was unplugged from the splitter.😔
Temps are stable and the frequency remains set.....I still see my voltage drop from 1.34 to 1.26 during Prime 95load which of course cuases it to spit errors.
All I set Bios to was all core 50, XMP 2, manual voltage 1.345.
Everything else in bios is default.
Bios version is current 0506.
Isnt manual voltage fixed? P95 hits 100-103c about 1 minute into small fft....but that shouldnt cause a voltage drop right?

HiVizMan
Level 40
Cheers for the response mate. Trust me when I say it happens to all of us, Hell I have run my system with the cooler fan header off and pulled my hair out trying to figure out why temps had shot up...

Set LLC to 6 and adjust your power plan in OS to performance.

VDroop is not a bad thing by the way it is a design feature built into CPU's by Intel and does important work.
To help us help you - please provide as much information about your system and the problem as possible.

HiVizMan wrote:
Cheers for the response mate. Trust me when I say it happens to all of us, Hell I have run my system with the cooler fan header off and pulled my hair out trying to figure out why temps had shot up...

Set LLC to 6 and adjust your power plan in OS to performance.

VDroop is not a bad thing by the way it is a design feature built into CPU's by Intel and does important work.


Load line is set to 7 by default ( AI optomized). I still dont understand this insane mess. I have the Vcore set to 1.314 Manual ( No adaptive).
At idle at desktop CPUZ shows Vcore @ 1.314 solid ( just like I set in Bios) ( bios recommends 1.28v for AVX stability @ 5ghz).
The moment I hit small FFT I get errors before temps even hit 70c and sure enough CPUz shows Vcore not at 1.314 but 1.277. So in order to get 1.314 at load I have to set my Vcore to like 1.36 or 1.38 so Im actually drawing over 1.29 volts....however temps get well into the 100's then.
How do I get my Vcore to just sit where I tell it to during load???

HiVizMan
Level 40
Start again mate.

Wipe that first effort away and start afresh. F5 yes and save.

Then at defaults boot to Windows and change your power plan to performance. If you have been using any OC software in OS tell me, and or remove it.

Then start with small steps.

In BIOS manually set LLC6
Manually set sync all cores and select your target multi
Same again with the vcore set to manual at enter what your CPU is happy with.

Leave it at that for now. F10 to save and exit.

Run Realbench please and only have one application open at a time that is polling the different sensors of the motherboard and CPU. Lets nail this OC yeah?
To help us help you - please provide as much information about your system and the problem as possible.