So like 45mhz I lose for no good reason? Nice :\
You see, I'll never reach those temps. Even at full load the card stays under 70c because of the good cooling, and the only way it's getting any hotter is if more power is given, which isn't given by default because already too hot according to that stupid boost. This really doesn't make any sense to me. The card is being boosted, but then throttled (sorry, "adjusted") simultaneously right from the start of being boosted. It would make sense if this happened at a more reasonable point, like 70, or maybe even 60, but to start pulling the punches at 40 is crazy. Why is this even a thing if its not the throttle? Isn't this just doing what the throttle is meant to do, only much earlier?

True thermal throttling would pull like 300+ mhz, so 45 MHz or 75 MHz is a drop in the bucket.

Of course it's small compared to what would be taken when temps hit the 'true' throttle, but what is that 45mhz even doing? By the same logic that little of difference isn't saving me anything for temps, so what's even the point? It seems like a complete waste of speed.

The logic is that is just how nvidia gpu boost 4.0 operates. Nothing we can do about it expect for lower our temps. My 2080 ti kingpin only loses 15 MHz as it stays around 40c. 

I can stay around 40c when playing some older games, but newer games are more demanding and push me up to 55-60.

My old 1080 ran up to 80 before I cooled it down to 60 on water, and it ran 2100mhz constantly. Since when did Nvidia get scared of temps? :(

do you go up the stairs using every single step or jumping 4-5 at a time, just because you can? 
 
reducing boost clocks is NOT thermal throttling, just reduction (auto) oc clock.
at some point the chip has to drop the clock to stay within given limits, and Nv wants this to happen gradually, not a sudden drop because temps spiked for a sec.
do you really think they will have happy owners, if ppl see sudden fps drops in-game, just so "you" can have another 10-50 mhz all the time?
(and a 20xx series isnt a 10xx series, so just because the 1080 did something, doesnt mean the next gen will as well)
 
next time get a card with a complete block.
my 2080S runs 28C on idle/54C under full load, and thats with the heat from my 3700x (sits before the gpu) on a 280 AIO with a throttled 18 gal/h pump.
 
 
 

My card no longer goes any higher than 55c or less under full load, and it idles at a cool 25c. Typically between 40 and 50 when gaming. I have even obtained consistent speeds over 2000mhz. 2050 under moderate loads, and no less than 2010 under full load and maxed volts. A full block isn't really necessary. I replaced that annoyingly loud generic aio with my old corsair I was using on the old card, and I replaced the stock rad fans with two noctuas. This all allows my card to maintain higher volts and constant speeds.

Well, the point is it is not "card reduces clocks if its hot" and it is not thermal throttling - it is "card is boosting a tiny little more if its cold", so it is boost implementation based on temps (colder silicon is - lower voltage it requires for same frequency). You shouldn't upset about that because first of all - it is actually quite small (as it said - about 15mhz addition, so for example my card can run up to 2130mhz under 40C or up to 2100mhz under 60C/2085mhz under 65C, difference is less than 2% and is not that significant. But to achieve that <40C I need to run all fans at full speed plus set ambient temp to 18C or less, while in silent mode (<900rpm) I still have a great performance and my pc is completely silent.

Your opinion of performance is personally subjective, and based entirely on what you run with your card. The percentage of Hertz difference is also not a 1:1 comparison to performance.

My evaluation of card performance is based on the average numbers recorded by others, wherein I find myself falling below that average. For a custom built card one should expect to fall above the average.

The only other explanation would be that I'm too honest about my numbers, and most others inflate their numbers for their e-peen.

Jiberish001
The percentage of Hertz difference is also not a 1:1 comparison to performance.

Well, actual performance is quite close to 1:1 in absolute, but yeah, performance improvement is even less than that. So this boost is not a big deal. Even if we are talking about something like 36C vs 86C - its like additional 90mhz to 1.9-2.1ghz clocks. Not worth in regular gaming at all. So you should not be upset about that.
 

Jiberish001
My evaluation of card performance is based on the average numbers recorded by others, wherein I find myself falling below that average. 

And what are you actually looking at?
AFAIK average 2080ti scores 14500 in 3dmark time spy, while ftw3 hybrid even without overclock and with default powerlimit should score around 15500. And temperatures doesn't change score significantly. 
 
Recently there was a topic about 2080S and performance lost (score dropped frop 11500 to 9000, same for games, temperature dropped too), issue was fixed with reseting px1 settings: for some reasons it showed powerlimit as 100%, but actual limit was much lower.