Okay so back in January, before installing the 3080 I ran my 2080 Super in Uniengine Heaven and got:
 
Average FPS: 136.7
Score: 3443
Temps: 69c
 
After using DDU and installing the 3080 drivers/Card and then running it again with exact settings:
 

 
Awesome expected difference right?   Then a few months later the pump is dying.  I RMA it.  I receive another exact same card.    Today I go through the same steps of DDU, downloading latest 3080 drivers and run uniengine heaven again. To my horror this is what I get,  same results multiple times:
 

 
%40 slower than the first 3080 XC3?   Basically a couple frames faster than the 2080 Super?
Booted up Halo Infinite and played for a while, and again getting the same performance I got on my 2080 Super   (Had it back in waiting for the RMA).      This time there is lots of stuttering and freezing going on.  Sometimes the freezing lasts about 5 seconds before gameplay resumes.
 
Things I have checked:
 
- Card is %100 seated properly.
- Both power cables are fully plugged in and connected
- Nvidia control panel is set to Prefer Maximum Performance
- Did DDU 2 and reinstalled drivers 2 more times.
- Fired up latest PX1 and reset the card to default settings.
- Ran MSI in the background to make sure it matched the in game benchmarks,  temps were the same as my last card, maxing out around 48c
 
I get this feeling there's some dipstick mistake my brain has made but I just can't think right now as it's approaching midnight and it was a long sleepless week.  Thanks for any suggestions.

Hoggle
First screenshot shows numbers without 8XAA Fullscreen enabled. You could try running the RMA replacement without 8XAA and see if the numbers are about the same.

Yep. I just ran it and
FPS 156.7
Score 3946
on a 3080ti.
at 1440p with all the settings maxed.
 
So he's about right where he should be with a 3080. I ran that one at 2100mhz with a bit of an undervolt.

HeavyHemi

Hoggle
First screenshot shows numbers without 8XAA Fullscreen enabled. You could try running the RMA replacement without 8XAA and see if the numbers are about the same.

Yep. I just ran it and
FPS 156.7
Score 3946
on a 3080ti.
at 1440p with all the settings maxed.
 
So he's about right where he should be with a 3080. I ran that one at 2100mhz with a bit of an undervolt.

By maxed do you mean with 8x on?

 
Problem is solved thanks to Hoggle and Hemi.  I do find it interesting that AA in this benchmark takes about %40 performance hit, whereas when I test out AA in other modern games the hit is probably 10 to 15%