Anyhow...Conductonaut is astonishingly easy to apply if you start with a tiny amount and just 'scrub' it into both surfaces with the applicator until you have evenly wetted the contact area with a thin layer. The most hazardous period is ensuring you don't blast a blob out of the syringe and ensuring your working surface is flat and level. Redoing a H100i V2 and 1080 Ti took all of an hour. As Broadwell-E is soldered and the GPU is bare die, I wasn't expecting a lot. More of an experiment and as a long term (it never needs replacing) thing.   Both the CPU and GPU saw a reduction in load temps of about 4C in extended gaming sessions. 

clamatowas
I just COnductonaut on my 1080ti. I used Nail-polish cover to coat all the components and made sure I used very little of the material making sure there was no sigh of pooling.

https://imgur.com/vxjQFYC Stock / Min 30c - Max 75.  Max fan Speed 61%
 
https://imgur.com/5Bwehqw after the Liquid Metal / Min 26c - Max 67.  Max fan Speed 49%

Awesome results!

I used liquid metal on the die when I delid my 7700k last year... but used GCExtreme on the Kraken x62 waterblock and for the 1080 Ti waterblock when I disassembled my FE card and installed the hybrid kit.
 
Everything I read said liquid metal was a big risk outside of what I used it for... so I followed the advice.   A year later temps are the same at idle (20's C on CPU and GPU) and at load (60-70C CPU and 40's GPU) as they were on Day 1 so I'm happy.
 

I know this thread is a bit old but here is a thought I had when looking over it. Everyone was concerned about the LM getting onto the PCB and shorting the cards out. With the heat sinks being on the lower side of the cards wouldn't gravity more likely have any LM that was pushed out move onto the heat sink and not up onto the baord?
Besides the same precautions with nail polish that are used on CPU's should be taken

GGTV-Jon
I know this thread is a bit old but here is a thought I had when looking over it. Everyone was concerned about the LM getting onto the PCB and shorting the cards out. With the heat sinks being on the lower side of the cards wouldn't gravity more likely have any LM that was pushed out move onto the heat sink and not up onto the baord?
Besides the same precautions with nail polish that are used on CPU's should be taken

Often times the failure occurs when too much LM is applied and when the heatsink is sandwiched back on, the LM can be pushed out onto capacitors/components. Keep in mind, LM will eat away at the die over time and nail polish or any materials, other than TIM, applied to the chip/PCB will void the warranty.

a lot of talk about conductonaut here, everyone is aware that coollaboratory liquid pro is superior, right? lol

also, i am correct in understanding that replacing the tim on the gpu DOES NOT void the warranty as long as it doesn't break anything?

coollaboratory pro has an 80W/mk rating. the 32.6 floating around is an error. it's not a huge difference really when you get up that high, the benefits start to level of, but it is better for pretty much the same price.

Fearsome Kitty

also, i am correct in understanding that replacing the tim on the gpu DOES NOT void the warranty as long as it doesn't break anything?

 
Correct EVGA allows you to remove the heat sink / change the TIM as long as you cause No Damage to the card
 
Liquid cooling is better an safer than any conductive TIM

Fearsome Kitty
http://www.coollaboratory.com/product/coollaboratory-liquid-pro/
 
unless they are falsely advertising, then it is 80W/mk. i would be very disappointed is these numbers on their site were not accurate. I have contaced them about this to confirm it.
 
and thanks, I did not know that I could replace the tim. not sure if i need to tho since mine hasn't gotten above 59 degrees yet lol.

That is really weird because Anandtech has been around a long time and are not known for being that wrong, especially citing the CEO. We could kill two birds with one stone if Anandtech is wrong.

Fearsome Kitty
when i was researching the liquid metals that was one of the places i looked at. it was the only place that listed liquid pro at 32.6, besides other places citing that test. I decided to go with coollaboratory's claimed rating rather than what was posted elsewhere in the end and had a 20 degree drop in cpu temps. tbh tho there are diminishing gains, even if i were to put a layer of diamond between the die and ihs it would barely cool any better than liquid pro or conductonaut, there just isn't any real world difference to be had once you get as far as those pastes. between the ihs and your cooler it makes even less of a difference with little to be gained when you go beyond only 30 W/mk, so i just used kryonaut, which is rated at like 12 or 13.

Personally I've found LM to be more consistent in application as long as you take a bit of time to scrub a nice thin shiny layer on each surface. It doesn't degrade is the other plus. For GPU's it's also true that I saw only a 2C or so drop in temps going from the stock thermal paste to the LM, and a similar amount on my CPU.  I think we agree that if you're starting with a proper application of thermal paste and a soldered IHS or bare die, in the case of a GPU, your gains are modest switching to LM.

Fearsome Kitty
depends on which side of the ihs you are applying it to on a cpu, on the outside, there is basically no gain versus a paste, but between the die and ihs there are massive gains compared to a regular paste.

I have been looking around more and there are other places listing liquid pro at 80, seems that one test by that guy on anandtech is the only place listing it at 32.6. Still waiting on an email response from coollaboratory.

It only depends if you don't have a soldered IHS. I've been at this while. 

Fearsome Kitty
oh yeah, if it's soldered then you set c: I like that they have stopped soldering them, gives you the option to take the ihs off and change tim or just leave the ihs off and do direct die cooling.
 
I agree that it isn't that bad to apply liquid metal if you are just painting a small amount of it on instead of using a glob and mushing it around like paste. I normally just put enough of it on that the whole die is shiny and then rub the excess on the other surface, when you put the two pieces together there shouldn't be enough there for any of it to get pushed out, unlike what some people do in youtube videos. unlike thermal paste, it really doesn't matter if you used too little as it will melt and even out. it would be really difficult to actually use to little for it to work as long as you made the entire die shiny.

Yes, as I said in my application tips in post #61 above. Sorry, but I can't get used to people repeating things back to me I've just said.

I have had liquid metal on my 1080ti FTW3 for over 7 months now. Temps out of the box are great but with the liquid metal at a room on 18c Ambient my temps were great. The only game EVER to have made my card go beyond 60c on 50-70% fan is Final Fantasy 15 Windows edition (this game obliterates my 1080ti haha) 
 

Does this really void the warranty? I also have such stains (a bit less though) on a H60 where I used Arctic Silver between the CPU and the H60. I mean the stain isn't affecting the cooling that much otherwise you'd have a big problem now no matter which TIM you apply?

the_Scarlet_one

Fitfit
What can actually damage the GPU chip when using that liquid metal?
Is there a way to find the components somewhere?

The liquid metal could make contact with other components on the board, and short them.

Some liquid metal, like liquid ultra, will also dry out and harden the cooler to the core, which risks cracking the core. Supposedly conductonaut does not dry out.

Have you experienced this happening yourself?

Gallium is extremely soft and have a melting temperature of only 29°C .. I don't see it "cracking a die" it's impossible.

I have clu and clp on almost every CPU and GPU I've ever had since it came out. Never experienced an issue.

I also thought the ftw3 block looked kinda weird, but evga says it's copper and nickle, and not aluminium so.. shouldn't be an issue aslong as you put some nail polish around it. Even then that's a precaution that's good, but not necessary.

I've had clp dry out and harden on a 3930k with an ekwb after 6 almost 7 years 20h/day operation. When I changed the liquid in the loop it flogged and clogged so I figured I might aswell put a new block on.
There was zero issue separating the parts, but the clp was hard and impossible to remove. I just applied some fresh clp on it which makes it soft again, and removed it with ease. Then a new coat of CLP and put it back together.

I have it on a 780 for a couple of year until I changed the waterblock back to the aircooler, no issue there. Had not even hardend.

A delidded 4770 ... w/o ihs .. clu inbetween for 5 years .. not a problem.

I don't doubt people can experience issues with aaanything ..just that I have not seen any issue myself and I've used it alot.

Aluminium is the poor man's copper, and should not be used in heatsinks or cables.

The card I have in my gaming rig now is a 1080ti ftw3 eg .. which I believe could have deserved a kpe cooler on it, since they were basically the same price. Now that I know the fins are aluminium, I'm really disappointed. EVGA is not a brand that cuts corners like this!

I guess I'll just throw a block on it now instead. Not that it will do much on performance, but I won't be able to look at the card now that I know it's dirty.

I almost want to throw the entire thing in the trash now, and get a rtx straight from Nvidia. I'm really disappointed atm.