cancel
Showing results for 
Search instead for 
Did you mean: 

Maximus Formula XI Z390 - VDroop

Wih__Glah
Level 7
Hi!

I am seeing a significant VDroop on the VCore of my 9900k.

I set the Vcore manually to 1.22 in the BIOS. During normal activities it drops to 1.18v and during AVX it drops to 1.14v

All cores at 5GHz, AVX offset 1. LLC set to 6,7 or 8 makes no difference.

Power limits have been maxed as per Debaur's guide on 8700K and Intel ETU states I am not power throttling.

Any ideas?
219 Views
29 REPLIES 29

Shamino
Moderator
Elmor has explained the new truer method of reading voltage of the M XI's here:
https://www.overclock.net/forum/27686004-post2664.html

Shamino wrote:
Elmor has explained the new truer method of reading voltage of the M XI's here:
https://www.overclock.net/forum/27686004-post2664.html


It's always been known that software reads are not wholly accurate. That's why to a large part I work with offsets and then ma mange the voltage to find stability rather then set a voltage I think should work in a certain way and then go from there.

It's also why when I have set an adaptive voltage I have quite a low llc value compared to many overclocking but have used the boards cpu current setting which allows more power delivery and an offset to balance things out as it gives me finer steps of voltage than just trying to use vcore adjustments and llc.

It's why I'm stable with voltage readings similar to what I'm going after int he bios and as Elmor suggests am working off a lower voltage read than bios setting.


Hasn't it been the standing advice for years that software doesn't read properly?

Hey did anyone find a fix for this?

My xi formula is the same. Set 1.33v bios and get 1.27v under load llc7

Jaz11 wrote:
Hey did anyone find a fix for this?

My xi formula is the same. Set 1.33v bios and get 1.27v under load llc7


There is no fix as this is normal behavior for the bird. LLC 8 is needed for no droop. *Just be careful of the potential overshoot on voltage transition. Some droop is a good idea for daily systems.*

Zammin
Level 9
So it's not just me experiencing this. Same deal when I was playing around with manual voltage on 9900k and XI Formula, even with LLC7 there is notable Vdroop. I'll have to try adaptive and setting SVID behavior to best case and see if anything changes.

EDIT: Whoops, didn't see the posts above. I'll have a read of Elmors post on OCN first.

tostitobandito
Level 7
Yep, same on my Hero XI which has the same VRM. I mean it works fine, but the amount of droop at LLC7 is a little frustrating. I'm stable at 5.1GHz at 1.305v adaptive with AC & DC LL set to .01. Under max load my vcore reading drops to around 1.25v or slightly over. Are we supposed to use LLC8 and just set way lower voltages?

Edit: Looks like Elmor is saying over on overclock.net that the M11 is just measuring vcore more accurately and closer to the CPU than most other boards. I've asked him to clarify since I don't think the post he made earlier was entirely clear about that. It explained the new way Asus measuers vcore, but it didn't really say that every other board out there does it differently and will show way higher vcore readings under load, which I think is what he just implied.

https://www.overclock.net/forum/5-intel-cpus/1714622-9900k-large-vdroop-load-voltage-main-1-watch.ht...


So assuming that is correct, if you were actually able to measure voltage at the CPU with a meter you'd see vdroop similar to the M11 boards on most/all competing boards. At least in the same ballpark, subject to variations for VRM design and so forth.

tostitobandito wrote:

So assuming that is correct, if you were actually able to measure voltage at the CPU with a meter you'd see vdroop similar to the M11 boards on most/all competing boards. At least in the same ballpark, subject to variations for VRM design and so forth.



Yep, that is what it means.

So essentially other boards just arent reading their actual load vcore as accurate as the asus?

Mine seems extreme. 1.34v llc7 bios and droops to 1.27...

Jaz11 wrote:
So essentially other boards just arent reading their actual load vcore as accurate as the asus?

Mine seems extreme. 1.34v llc7 bios and droops to 1.27...


Currently, no other vendor utilizes differential sensing directly from the pins, so the reading on will be less accurate. All else being equal, on a board that uses conventional sensing, the sag would be more masked (the actual magnitude depends on the sense location).

So asus vrm reads the voltage differently and more accurate showing the droop?