I'll never use anything but Conductonaut again, ever..Never ever..Never ever ever never..The stuff is amazing..Thermal Conductivity of a whopping 73 W/mk.
 
How Liquid Metal Affects Copper, Nickel, and Aluminum (Corrosion Test)
 
If I was to go with paste, it would be Kryonaut.

I can tell you is in my experience with MX-4 is that it just won't pump out and is the best set it and forget it for years paste.  My Alienware 17r4 with a 7820HK and GTX 1080 had the well documented thermal problems with the stock Dell application as did my 13r3 wiht a 770HQ and GTX 1060.  Popular thing for these models is to Conductonaut or Kyronaut them.  Not a fan of LM conceptually, I opted for the latter.  Temps were massively improved for both but slowly over a month or two would worsen to the point of thermal throttling.  After several Kyronaut repastes I repasted both with my old standby of MX-4 and although it didn't have the same absolute performance of Kyronaut it was within a degree or two for any CPU core at Kyronaut's best readout and for almost a year, has not been reapplied or deviated from day 1 performance.  Take away here is I don't recommend Kyronaut for direct die CPU applications given my experience.
 
I did use Kyronaut on my 2080 Ti FTW3 Ultra when I converted first to hybrid and now as it has been since Feb (iirc) under a Hydro Copper.  Started off with a delta of 11c vs coolant and it has since closed down to 8c at max loading, the thermal density of the TU102 and of GPUs doesn't seem to be a problem for Kyronaut.
 
When I assembled my 2700x/Crosshair VII Hero build in July 2018 I built it with the intention of doing a Zen 2 upgrade which was done in July 2019.  I ran that 2700x quite hard.  Sometimes on a manual 4.3Ghz overclock which would see folding @ home CPU workunits eating damn near 200w, sometimes I ran in PBO with a negative core voltage offset where it would max out at 150w SVI2 sustained consumption.  Temps never varied in that application over the year I had the 2700x installed and were consistent for a given loading scenario.  I used Kyronaut there and did so again when I dropped the 3900x in.  It just doesn't seem to pump out as fast as my intel laptop CPUs over multiple applications... so long as there is a good amount of surface area that an IHS provides so I would argue that it will be just fine for you long term in the application you propose.  How much of an improvement over the EVGA pre applied paste?  I have serious doubts it would approach 10c unless a lot was smudged out during installation and for a perfect pre app by EVGA and install by you I have real doubts it would be 5c difference in Kyronaut's favor.  My expectation assuming a good TIM interface from the CLC kit and a good clean and repaste would be 0c > net c > 5c when using Kyronaut or MX-4 where I have always seen Kyronaut perform at about 1c better than MX-4.  Since it seems you are looking more for absolute performance than multi year consistency of performance I say give Kyronaut a shot if you have it on hand and if you don't but you do MX-4 just use that.

bobmitch

bob16314
I'll never use anything but Conductonaut again, ever..Never ever..Never ever ever never..The stuff is amazing..Thermal Conductivity of a whopping 73 W/mk.
 
How Liquid Metal Affects Copper, Nickel, and Aluminum (Corrosion Test)
 
If I was to go with paste, it would be Kryonaut.

Is Conductonaut good inbetween the AIO and CPU?  I read that too...wasn't sure what the cover of my CPU is...but said don't use with aluminum...

iirc Intel IHSes are nickel plated and s/b just fine.

 Thermal Grizzly Kryonaut is the one I've used in the past and would not think twice to use it again.

i've been using the TF8 on everything, CPU's and GPU's and its been 0 issues and my best OC's recorded with it.