Worth repaste/LM?

[Cris90]

Hi I’m upgrading my system but keeping my 3080ti evga ultra hydro. 
 

I have a custom water cooler set up very overbuilt, 2x360 rads, and 9 fans, for my gpu and 3900x. 
 

im going to a 7500x, new MB and memory, so everything has to come apart for the water cooling anyways. 
 

currently my cpu is my bottle neck, is it worth me repasting/using Liquid Metal on my 3080ti (currently it doesn’t over heat) but not sure with the new set up if it’ll be pushed harder as it’ll probably be the bottle neck now. 

[OhioYJ]

Liquid metal is a hassle in my opinion. There is a technique to applying it. You can have quite a bit more variation in temperatures between a "good" application and a "bad" application. Then on top of that you have to deal with all the other issues liquid metal brings. 

 

If you aren't going for every last bit of performance (IE, pushing the limits of overclocking this card) just use a quality thermal paste. 

[Agall]

20 minutes ago, Cris90 said:

is it worth me repasting/using Liquid Metal on my 3080ti

I'd say no, unless you used some bad paste to start. I've come across so many systems where the paste is +10 years old and still in great shape. The more sealed it is, the better it seems to do, a graphics card waterblock being fairly sealed to the die.

 

If you want something cool, look at a kryosheet or a honeywell ptm7950. Liquid metal works as well obviously, I just don't think its overall worth the hassle on something already waterblocked. The only reason I used liquid metal on my 9800x3D mITX setup is because I had a spare tube laying around that I got when doing a direct-die mod for my 7950x3D.

[PDifolco]

It won't look make any difference if the card is already not hot

If you want to do it anyway you should use PTM7950, way easier than LM and last longer than normal paste

[podkall]

12 minutes ago, Agall said:

The more sealed it is, the better it seems to do, a graphics card waterblock being fairly sealed to the die.

I'm pretty sure the range it increases and decreases in temp also determines the wear of the thermal paste

[Agall]

3 minutes ago, podkall said:

I'm pretty sure the range it increases and decreases in temp also determines the wear of the thermal paste

I agree, there's a few variables, but especially with GPUs, pump out is something I forgot to mention. That's where a kryosheet or ptm7950 comes in.

 

@Cris90 Pump out on GPUs makes liquid metal to me not worth it, since eventually, after enough expansion and contraction of the die, you'll get enough gaps to cause issues with liquid metal. Its far more of a problem on GPUs than CPUs, especially lidded CPUs.