1.17 appears to work the same way.

If my understanding of the functionality of that setting is correct than I would say this is a bug... EVGA clearly states in the BIOS that "Vcore offset voltage: Final voltage equals to sum of (Target + Offset). EG 1.2 + 200mv = 1.4v." Sine this excepts negative values in the range displayed in the BIOS than I dont think that its unreasonable to say that it should be 1.2 - 100mv = 1.1v certainly not 1.3v... Really at 1.2v I would suspect that it would do nothing. 

I am going to reach out to EVGA support tonight and confirm that my understanding of the settings is correct and report the bug. Hopefully this gets corrected soon... 

jasoncodispoti
If my understanding of the functionality of that setting is correct than I would say this is a bug... EVGA clearly states in the BIOS that "Vcore offset voltage: Final voltage equals to sum of (Target + Offset). EG 1.2 + 200mv = 1.4v." Sine this excepts negative values in the range displayed in the BIOS than I dont think that its unreasonable to say that it should be 1.2 - 100mv = 1.1v certainly not 1.3v... Really at 1.2v I would suspect that it would do nothing. 

I have the Same Settings Same CPU Same Memory they both show different numbers but not a negative one.
I would say it is more CPU Silicon Lottery than a Bios Bug.
Maybe it is just reading it incorrectly or is by design.
You are also using Adaptive and Setting the Mesh Ratio to 30 and not showing the rest of your Bios OC Settings.
I use Override and stopped using Adaptive after Bios 1.09.
What CPU are you running?
Are you using Two EPS Cables or Only One?

jasoncodispoti
I am not sure how silicon lottery plays into this? Please explain. If I set the target vcore to 1.2v and the offset voltage to +0 voltage does not exceed 1.234v even under load. However if I set vcore to 1.2v and the offset voltage to -100 voltage increases to 1.332v, that is almost a perfect 100mv increase. When it should have decreased voltage by 100mv so it should have been 1.2v - 100mv = 1.1v not 1.332v. 

Just go by whatever the EVGA Forum Moderator tells you, so if they say it looks a Bug in the Bios then it must be.

jasoncodispoti
I am not saying that you are wrong I am just trying to understand what you are saying... either way I will call EVGA tonight and get an answer. 
 
Thanks! 

Not saying that I am right as well.
But at least I can test this on my own hardware being that I do own this Motherboard.
Keep us posted of what EVGA finds for you.

bcavnaugh

Sajin
Looks like a bug to me.

Your Evidence?

LOL. Are you serious? He entered a negative value, and it increased his vcore. That is clearly a bug. That isn't how it's suppose to work.

https://forums.evga.com/FindPost/2905393
 

TiN_EE
Mostly 1.2V min Vcore voltage is due to Intel CPU's VID programming range. When you have Adaptive Vcore it's not just simple X.X volts for VRM and CPU running that specific voltage at all times, but rather complex inter-operation between motherboard VRM, physical layout and own CPU power delivery regulator (FIVR). On software side there are three things that control voltages and power - BIOS code, PCU firmware (inside of CPU) and FIVR settings. 
 
In theory BIOS can program any voltage, below 1.2V including, but if that voltage is below CPU's own VID point for current running frequency then CPU will ignore setting you program and will settle to own lowest voltage. Also I don't see much reason why you want lower voltage, even if CPU could run it in some cases, undervolting would cause stability issues on heavy-loads/turbo clocks. There is not much safe margin between CPU VID and stable clocks to make a large difference. Great tool for testing this would be Prime95 (non-AVX older version) and AVX newer version as it stresses chip pretty hard.
 
Adaptive Vcore also does NOT mean cores will be running at 1.2V (or more) all the time, with default and even OC with adaptive Vcore setting voltage will be adjusted according to CPU loading, going down to 0.8V or lower, depends on CPU. To say differently: if you running adaptive and set voltage below what specific CPU have predefined from factory - voltage will be set by CPU, not MB BIOS. So we removed option to set below 1.2V at BIOS setup as it would have no effect.
 
CPU will only run fixed voltage if you use Vcore override setting and set it over what's CPU default voltage level. 
If you get good CPU (with plenty of voltage margin), I can recommend to try with Vdroop settings, perhaps there is specific configuration that can help you to maintain stability and let CPU run at lower voltages. Other point - you want to keep hottest die temperatures below 83C to maintain best performance. 

Lowest target voltage should be less than 1.2 V instead of limited to 1.2 V...……..

I just conducted another test which I feel continues to confirm that this makes no since... 
 
Running, 
vcore @ 1.2v and voltage offset at +0mv I get 1.2v (as reported by the BIOS on vcore)
vcore @ 1.2v and voltage offset at -100mv I get 1.3v (as reported by the BIOS on vcore)
vcore @ 1.2v and voltage offset at -200mv I get 1.4v (as reported by the BIOS on vcore)
 
This just makes no since to me... why with all of the settings configured exactly the same for each test would the CPU request more voltage as a negative offset is put in place. 

EVGATech_LeeM
 
I had our BIOS team look into this last night, and they were unable to reproduce this issue. 
 
Can you let us know which CPU you're using, and can you provide us with your BIOS profile?  I'll forward these to our MB Team.

I appreciate everyone looking into this, here is a bunch of details, hope it helps! 
 
PC Build Specifications: 
Motherboard, EVGA X299 Dark BIOS v1.17
CPU, Intel Core i7-9800x
Memory, Corsair Dominator Platinum DDR4 3200 4 DIMMs Total of 32GB
Hard Drive, Samsung 970 PRO 512GB NVMe PCIe M.2 (installed in PM1)
Power Supply, EVGA SuperNOVA 1000w P2
Graphics Card, EVGA RTX 2080ti FTW3 Ultra Hydro Copper (installed in PE1)
 
OC Settings: 
CPU Multiplier Control, Manual - RatioLimit
CPU Multiplier Setting, 45 (all cores)
Mesh Ratio, 30
AVX2 Negative Offset, 4
AVX2 Negative Offset, 4
BCLK Frequency Settings, 100.00
Extreme Cooling Mode, Disabled
Extreme Voltage Mode, Disabled
CPU VIN, 1.800
Vcore, Adaptive
Target Voltage, 1.200
Offset Voltage, +0
Vmesh, Override
Target Voltage, 1.100
VIN Vdroop, Enabled
VSA, Auto
VCCIO, Auto
PCH 1.0v, Auto
Uncore Voltage Offset, Auto
PLLtrum Offset, +0
MC PLLtrim Offset, +0
 
Description of Issue: 
With all of the above settings configured in the BIOS and a vcore voltage of 1.200v and offset voltage of +0 the BIOS reports CPU voltage at 1.250v. That voltage is reported within the BIOS screen and within the built in EVGA Stress Test in the BIOS. However if I set a negative voltage offset with all of the settings remaining exactly the same, so -100mv (voltage offset). The BIOS reports CPU voltage at 1.347v roughly a 100mv increase over the previous reported settings. This voltage is also confirmed within the built in EVGA Stress Test in the BIOS.
 
I am well aware of the post from Tin_EE regarding vcore voltage being set at 1.2v and attempting to set it to a lower value is somewhat pointless... However I dont feel that applies in this situation as the would technically mean that I would basically see no change in voltage when setting a -100mv voltage offset and vcore of 1.2v, certainly should see an increase in voltage though. I have attached two BIOS profiles to this post... One labeled Offset voltage +0, this is just my standard normal settings that result in a typical voltage of 1.2v give or take. The other is labeled Offset Voltage -100 this is the one that results in a typical voltage of 1.3v. 
 
Could not add the files to the post so here is two links to get them from Google Drive, 
https://drive.google.com/open?id=1pkRqpG-QQAVj9zlCZ5qNgokBFo5wqhap
https://drive.google.com/open?id=1VVvRgX5RMt1pSOt7i2CegYhylxOQnowf 
 
Hope this information helps and thanks for looking into this! 
 

The fix in the beta BIOS was added to official releases, starting with v1.18.

It seems to be very hit or miss.
To the EVGA crew, the way adaptive voltage is supposed to work is by taking away or adding to the VIDS that intel has programmed into the chip.
Now my method of settings the voltage when I do it on my asus boards is to set a multiplier in the BIOS but leave the vcore set to auto.
I will then boot into Windows and load the system with something like cinebench or something non AVX.
What I will then do is note down the core that requires the most voltage, so lets say for example core 4 is programmed at 1.320 at a 47 multipier, If I boot back into the bios I will set the target voltage to that 1.320 value and if I want to lower it I will subtract the desired amount using the offset.  You cannot take too much away or you risk running into BSOD's or the board not even booting but if I dial in a negative offset of 100mv once booted into windows and the core is loaded it will run exactly that 1.320 minus the 100mv I took away.
The EVGA BIOS does not seem to follow that method at all and instead does whatever it wants, and it's different depending on the multiplier you have selected.

I think what Other users and myself would like is whatever we set the offset too let it run that, if the end user sets the offset too far it just won't boot but the current way it works leaves the end user totally confused why it won't work vs if the machine just failed to post they would know they went too far.
Hell i've had my X299 dark do some weird stuff, I'd set a target of 1.3v with a negative 100mv offset and boot into windows and it be running 1.5v when it should have been 1.2(yes I know below intels spec but that's what I commanded not 1.5)

I am not sure if you got my last PM but I'd like to see other users see if they have the same issue I am having using a per core overclock vs a single multi overclock.
If I try a per core overclock the settings do not work and the CPU just locks itself at its all core default turbo speed which for my 10980xe is 3800mhz.