A Rather Random Thought

[FreedomEclipse]

I was just browsing Amazon for thermal paste and it just occurred to me that TPU never got into testing thermal pastes. Even Guru3D did to a lesser extent although they havent done one for a while but otherwise it seems to be the realm of Tomshardware and other tech sites.

I guess it was something not worth devoting resources into

 

[nomdeplume]

There was and appears to continue being a quite serious contingent of users tackling this. If memory serves correctly this site... well I hope you are prepared.



Will @lexluthermiester please come to the courtesy desk.

There is a gentlemaaaaaaIIIIIIGGHHHHHH AAAAAAAIGGGGGGGGJJJJJJJHHHHHHHIIIIIIIIIIIIIIIII AAAAAAIIIIIIIIIJJJJJJJJJJHHHHIIII n waiting here for you.

 

[Nike_486DX]

even same brand of thermal paste varies a bit from batch to batch, so a review wont be that useful.
Personally when i need the best performance, that can also do 2+ years on a single application, i go straight for conductonaut. Then if you need aluminum-friendly paste that doesnt pump out - there is hydronaut.

 

[cvaldes]

I was just browsing Amazon for thermal paste and it just occurred to me that TPU never got into testing thermal pastes. Even Guru3D did to a lesser extent although they havent done one for a while but otherwise it seems to be the realm of Tomshardware and other tech sites.

I'm not sure if there's much value in TPU putting resources into this endeavor. We've seen the results of Tom's Hardware on-going thermal interface testing and there are some very obvious takeaways.

1.) Liquid metal (as a category) performs better than conventional thermal pastes. It's also trickier to apply (especially for someone who isn't routinely applying thermal pastes) and is more expensive.
2.) Conventional thermal pastes are all within 3-4 °C of performance of each other. So the choice of which one doesn't matter that much.
3.) None of these test longevity of the thermal compound and perform over months/years. We saw Arctic withdraw MX-5 from the market due to extended performance issues and QC variability between batches.

Remember that for Joe Hobbyist, they might paste up a CPU or GPU once or twice and be done with it. So a pricey tube of some premium brand isn't going to matter versus a tub of industrial grade stuff that some PC repair shop owner might buy. The scale is totally different.

Much performance of the thermal compound is reliant on the skill of the person applying it. Someone who regularly applies thermal paste (testers, PC repair shop technicians, etc.) will have better, more consistent results than someone who does it once every 2-3 years. Just like decorating a cake. I assure you that if you don't have regular, long-term experience in working with cake frosting, you will not get professional grade wedding cake results.

Makes zero sense for TPU to start testing thermal compounds. Someone is already doing is and they've shown that it matters little which brand it is. It's more important that the paste is properly applied.

 

[Space Lynx]

thermal paste is dead.

just use thermal grizzly kryosheet, performs equal to even the best pastes. never need paste again.

thread closed. easy peasy. give me a hard challenge next time lads.

 

[oxrufiioxo]

thermal paste is dead.

just use thermal grizzly kryosheet, performs equal to even the best pastes. never need paste again.

thread closed. easy peasy. give me a hard challenge next time lads.

I do want to try those they seemed to work pretty well on the Asus 4090 super Lube edition after Hub removed the Liquid metal during their tear down.

 

[A Computer Guy]

thermal paste is dead.

just use thermal grizzly kryosheet, performs equal to even the best pastes. never need paste again.

thread closed. easy peasy. give me a hard challenge next time lads.

I wonder if they will come out with a sheet that wicks liquid metal. That would be much easier to apply and would help keep it from spilling out.

 

[Noci]

Nah, let the chip bakers use superconductor materials in their chips and we won't have any heat to deal with (superconductor at ambient temps, not near 0 Kelvin a scientists wet dream ).

Hey, you can't blame me for dreaming....

 

[Dr. Dro]

The biggest problem with testing TIM and cooling is that you need a controlled ambient with a controlled heat load to even make them comparable. With batch variance, it makes it really hard to assure people that what they are reading is what they will get when they buy. That plus if there's a really bad batch, it tends to spread on forums fast. Personally, I can say good things about the Thermal Grizzly Kryonaut Extreme, except for the price. I live in a very hot region and my i9-13900KS processor might as well be a fireball, but it's managed to keep it tame in conjunction with my id cooling 360mm AIO thus far.

Oh, speaking of which... @FreedomEclipse remember the IC Diamond thread? You linked this gem some months ago in some Alphacool thread if I'm not mistaken. Hahahah I still read this sometimes, epic forum fight, wish I had been around at the time... you old timers had it good

TechPowerUp! Official IC Diamond Test

I am guessing that is Aluminum Oxide residue from not completely cleaning off the previous TIM prior to ICD usage. Some sort of chemical reaction, much like a kidney stone!

www.techpowerup.com

 

[SK-1]

The community did the footwork on this one lol. I mean...Mayo FTW!!!

 

[Solaris17]

(as a category)

This. buy based on category. In the early 2000's it mattered a lot more. There were developments in all areas, from heat transfer to consistency, and few products on the market. Now this stuff is formulated and made by the barrel full and there are hundreds of companies and brands.

Categorically they are so close (within 2-3c) that if that variance matters to you or is the difference between stable or not, you have other issues and thermal paste choice should not be the focus of your energy.

 

[lexluthermiester]

thermal paste is dead.

just use thermal grizzly kryosheet, performs equal to even the best pastes. never need paste again.

thread closed. easy peasy. give me a hard challenge next time lads.

This is not bad advice. Thermal pads seem to be an excellent way to go.

I was just browsing Amazon for thermal paste and it just occurred to me that TPU never got into testing thermal pastes. Even Guru3D did to a lesser extent although they havent done one for a while but otherwise it seems to be the realm of Tomshardware and other tech sites.

I guess it was something not worth devoting resources into

I did some testing in recent years. If you search the threads for my username and the keywords " thermal paste " you'll find them.