The GPU will only use the amount of power that is needed for whatever it is doing... 
 
I have an EVGA RTX 2080ti FTW3 Ultra Gaming Hydro Copper with a power limit of 373w's. Just ran the Heaven Benchmark with stock GPU frequency and voltage + power slider at max, this was the results... 
 
MAX GPU Core Voltage >> 1.094
AVG GPU Core Voltage >> 1.029
MAX Power Usage >> 347 watts
AVG Power Usage >> 281 watts
MAX GPU Clock >> 1965MHz
AVG GPU Clock >> 1,878MHz
 
As you can see I did not hit max power usage, I am sure however if I ran the test with an overclock applied that it would come closer to the max power limit maybe even hit it. At the end of the day its a good thing to not be using all of your power budget... 

What perfcap reason do you see listed in GPU-Z's sensors tab?

Vrel ... 

What cooling solution are you using? And what is your max boost frequency when running the benchmark?

As far as why would people go with the higher power limit... My GPU exceeds the 320w limit under stock frequency. However my GPU is liquid cooled which makes a big difference.

LasseK1981
Vrel ... 

Well, there you go.  This means that you aren't power limited.  So, the BIOS mod did work for you.
 
So, crank up the voltage if you want it to boost higher.
 
From my understanding, Vrel means that it can not reliably ("stable") operate faster at the present voltage.  More voltage usually makes it able to be stable at higher frequency.

jasoncodispoti
I

dwexpress

LasseK1981
Vrel ... 

Vrel is voltage related is what is holding you back. 

As far as I know VREL is “voltage reliability” and from what I have seen I would not pay much attention to this limit.

Edit:
I generally ignore this as it’s true meaning is not really known. At least I have not seen anything from anyone “official”, maybe Tin could chime in on what that means. It implies a voltage limit, but it seems to always be active does not really make any since too me...

It's true meaning is that the card cannot boost any higher and remain stable at the max voltage the card is allowed.
 
Again, Nvidia hard locked the max voltage the card can go to to 1.094v. A hard mod to fool the card into thinking it is drawing less power than it really is would need to be done to go higher.
 
Example, my 1080Ti is locked at a max voltage of 1.094v. I am water cooling the card and my max temps under load never go above 35-38 degrees C. My card never hits the max voltage limit of 1.094v and maxes at 1.064v.
 
With overclock on the card, it will boost to 2050MHz all day long never hitting any limits in GPU-Z at 1080p resolution.
 
It will only hit power limit when I run resolution higher than 1080p or run Time Spy bench (which is a 1440 res benchmark) but still will not use more than 1.064v because the chip does not need more voltage at such a low load temp as the card is hitting the max power draw my BIOS allows (stock BIOS). Higher voltage would not improve the boost clock as 2050MHz is the max my chip can do and be stable.
 
In my case, more voltage would not net me any higher MHz frequency and would most likely degrade the performance of the card.
 
Nvidia boost is very smart, taking all factors (voltage, power, temp) into consideration to achieve a stable MHz frequency.
 
With Pascal and Turing (RTX) more voltage does not guaranty more MHz. These chips do not really like high voltages like older Nvidia chips did. They like lower temps.
 
The reality is, you do not want your card to hit the max power draw the BIOS allows no matter what power limit the BIOS you are using has as this will cause the card to step down the boost to keep within the performance cap of the card.
 
With Pascal and Turing, temp is king..........

LasseK1981
I am sorry I did forgot to mention the card IS in fact water cooled and runing within 60-63g under load.
 
But I noticed that its my cards max doing so it never go higher. My previous 2080 ti could take 2100 mhz whereas this one 2010 mhz if youre lucky. 1980 mhz is where it is mostly. So the reason why it dont pull 380 w is it cant simple as that. ITs hitting the ceiling before that. Thanks for all your input.
 
Lasse

While 60c is well within spec and certainly not "hot" it does seem high for a GPU under water... For comparison under heavy benchmarking/overclocking loads my GPU stays around 50c to 55c. From the testing that I have done GPU boost seems to really drop off at the 60c mark which would most likely explain your "low" clocks. Given that you have a founders card and it should be a better BIN than pretty much any GPU from a 3rd Party + water cooling + 380w BIOS I would think that you should be able to hit 2100MHz to 2150Mhz fairly easily when overclocking. Do you happen to know the temperature of your fluid? And the ambient air temperature?       

jonkrmr

jasoncodispoti

LasseK1981
I am sorry I did forgot to mention the card IS in fact water cooled and runing within 60-63g under load.
 
But I noticed that its my cards max doing so it never go higher. My previous 2080 ti could take 2100 mhz whereas this one 2010 mhz if youre lucky. 1980 mhz is where it is mostly. So the reason why it dont pull 380 w is it cant simple as that. ITs hitting the ceiling before that. Thanks for all your input.
 
Lasse

While 60c is well within spec and certainly not "hot" it does seem high for a GPU under water... For comparison under heavy benchmarking/overclocking loads my GPU stays around 50c to 55c. From the testing that I have done GPU boost seems to really drop off at the 60c mark which would most likely explain your "low" clocks. Given that you have a founders card and it should be a better BIN than pretty much any GPU from a 3rd Party + water cooling + 380w BIOS I would think that you should be able to hit 2100MHz to 2150Mhz fairly easily when overclocking. Do you happen to know the temperature of your fluid? And the ambient air temperature?       

Temp is too high for a sustained high boost. The card is notching down two bins in boost speed (2010 to 1980 = 30 MHz, each bin is 15 MHz) to keep the card's temp down.
 
It will continue to step down in 15 MHz increments as temp rises. This is why you cannot sustain a 2010 MHz boost and get anywhere close to the 380w power limit of the BIOS you are using.
 
Again and again, temps are king with the cards. If you cannot keep the temp under 45 C, you will not be able to get good over clock on it.
 
Granted the Turing cards run hotter than the Pascal cards do, but 60-63c is high for a water cooled card even under heavy load.
 
You need to check the contact of your water block on the card.
 
You might also need to evaluate your water cooling loop to make sure it is sufficient for the card.
 
With a proper water cooling loop and proper mount of the water block, your card should never go above 45ish C max under heavy load with an ambient air temp of 72 F.
 
Something is wrong.
 
High room temps where you are?

This  +1

Moving to Overclocking, Cooling and Benchmarking
 
Home » All Forums » [Computer Hardware/Software] » Overclocking, Cooling and Benchmarking
 
Your - 2080 Ti FE is not an EVGA card & does not belong in the EVGA 20 series section
 
Also your more likely to get help in the OC section --> with your custom BIOS

Hi again all.
 
Well I did in fact some more investigation and test. And it shows that I actually get a PWR Limit warning in GPU-Z when I try to break 2010 MHz..  At that point the card is runing 1.050-1.068v but it never and have never reached the 1.093v as I heard is a factory limit.
 
When the PWR limit comes and remember this is with the 380w BIOS. When the PWR LIMIT comes the card is around 320-340W / 2010 mhz.
 
The afterburner settings is
Core Voltage +100 max
Power limit 126%
Core +135  (crash eventually at 2010 MHz)  +130 stable at 1980-1995 Mhz
Memory +1000 works fine, no difference stock mem.
 
 
If I just have Core Voltage at 0 it will go to max 1.050v and around 1980 (with +130 core).
 
I am really trying to figure this out I think its a mystery.... 
 
HOW can I get a POWER LIMIT with a 380W bios when capping 330W?    How can I have a PWR LIMIT when Gpu vCore is 1.068v and not the rumored 1.093v    (IS the 380w bios = 1.068v max?)
 
Why does the GPU crash at 330w range due to insufficient power when apparently the BIOS should allow more draw and the PC should be capable of provide it?  2 x 150w PCI-E power cables + 75w PCI-E bus power = 375w 
 
 
Could there be some kind of hardware cutoff or could another bios help out?  I find the behavior strange. 
 
Temps are 60-65 Celsius range, CPU the same in bench and full load..
 
Lasse