https://xdevs.com/doc/_PC_HW/EVGA/E200/bios/2589L390.rom 520w vbios that doesn't void warranty.
https://xdevs.com/doc/_PC_HW/EVGA/E200/bios/rtxbiosxoc.zip 2000w vbios that will void your warranty.

sungandr860
Edit: :| E200 is the kingpin model

Correct. You said that you didn't mind voiding your warranty and accept the consequences.

See post #283 here: https://forums.evga.com/FindPost/3063949

It should work. Should give you a heck of a jump.

And if it doesn't work for some reason, make sure that you leave one BIOS unaltered just for fallback.

It should work. Only one way to find out. Make sure you leave one BIOS unaltered just so that you have a fallback.

So no boost. Have you tried setting your power management mode inside the nvidia control panel to prefer maximum performance? How about forcing max boost via msi afterburner voltage/frequency curve editor using the lock feature/boost lock inside precision x1?

sungandr860
Edit2: the boost lock works in the desktop, but as soon as Witcher 3 loads the gpu, it starts fluctuating between 1500-1650 again

Was this with the power management mode set to prefer maximum performance inside the nvidia control panel?
 
Edit: Just saw your edit.

https://xdevs.com/doc/_PC_HW/EVGA/E200/bios/2589K390.rom This one offers 520w too without voiding warranty.
 
Does that vbios also give you the same problem? That vbios is the kingpin oc vbios. The one you're currently using is the kingpin ln2 vbios.

Guess your only choice is to use the xoc vbios, but it will void your warranty. 

That can't be right or wouldn't be able to get over 800 watts from a 2080 ti

Yea the 150 figure I mentioned would be for crappy power supply units it’s more around 300 watts per 8 pin. I might flash it when I get home, if it’s anything like the YouTube video another 130 watts should provide a nice boost.

Yeah I know, to all there own for sure. I figure if ya want kingpin performance and options buy the kingpin.

Yeah a few of the kingpins would hit 2190 and a very few would hit 2205. The original AIO on the kingpin was very effective especially when adding better fans, even the custom loop was barely better. Once the chiller or other more extreme cooling was applied was when they went above and beyond by far

ty_ger07

sparetimepc
Even if you put the kingpin bios on your ftw3 your not going to get the same results because the PCB doesn't have the same components of the kingpin PCB. Your still going to be limited because of the cooling even if it's a custom loop. The only way of really benefiting from kingpin bios is using a chiller, ln2, or dice. The small gain from flashing the kingpin bios to the ftw3 is not enough to take the chance of bricking your card or voiding the warranty. An extra 30-40 watts is not a big enough improvement in gaming to visually notice your fps and make it worth it.

As I linked above, another user went from 378 watts to 590 watts using this BIOS on a non-Kingpin card. I think it will be noticeable. It's way may than 30 or 40 watts increase. 212 watts increase in that example.
As I said in the thread that I linked above, I think that it will cause premature GPU failure, but everyone is welcome to use their GPU the way that they wish. A long boring life, or a premature exciting death.

590 watts?!?! lol. nope. software readings are wrong. Let me show you how it really affects your card...
 


 
Note: Turn down sound so your ears don't get blasted.
 
 

Sajin
590 watts?!?! lol. nope. software readings are wrong.

Why?  That's the input to NVIDIA's GPU boost.  If NVIDIA's GPU boost operates based on information which is only 50% accurate, how does it operate at all?

Let me show you how it really affects your card...
 


 
Note: Turn down sound so your ears don't get blasted.

So, those software readings aren't wrong?  What's the difference?

Sajin
The readings on my card aren't software based. They are hardware based.

Well... that is semantics.  They are both software based.  Or both hardware based.  It depends on how you look at it.
The reading is a hardware reading, but the display on your GPU's screen utilizes software.  Or firmware.  Again, semantics when it comes to definition of software versus firmware.

GPU-Z readings are software based.

Yes, but they are also hardware based.  What is physically measured on the card utilizes hardware.  The exact same hardware which is used to display the reading on your GPU's screen.
GPU-Z is an abstraction, but it uses the official NVAPI which is what NVIDIA uses to control its boost algorithm, so it wouldn't be normal for it to be inaccurate by 50%.

I tried running gpu-z to see how it would do, and it seems like it was pretty spot on for me. Showed 333 for first vid, and 358 for second video. No idea how Cisco150 pulled 590w at 1.125v unless he ran furmark. 

That's what I also don't understand.