Best thermal compound?

[lerodemmy]

Are there any that stand out above the rest?

 

Also, how much should I really use?  I see some people use about the size of a grain of rice, and others make an X from corner to corner across the CPU.

[svmlegacy]

4 minutes ago, lerodemmy said:

Are there any that stand out above the rest?

 

Also, how much should I really use?  I see some people use about the size of a grain of rice, and others make an X from corner to corner across the CPU.

Depends on what you're doing. For general purpose stuff with an IHS, I use Arctic MX-4, as it's a long life formula that performs well.

 

The highest performance depends on how cold you're going.

[Skiiwee29]

Depends on what component its used on. Personally, Thermal Grizzly Kryonaut is pretty much the best out there for non liquid metal solutions, though all the top pastes will likely be within 1-2c of each other, which is basically margin of error and run to run variance as some pastes have a cure time and perform better after so many hours. 

 

As for application method, that again depends on the component used on. For mainstream CPUs, generally a pea sized dot in the middle and then mounting pressure spread of the cooler is adequate. However, with Ryzen its a little more complicated since the cores are not directly under center, so an X method or manual spreading is likely more optimal to ensure full coverage. HEDT processors such as Intel Extreme processors and AMD Threadrippers are a whole other different ballgame due to the sheer size of the IHS and again with AMD, the die placements since its not directly centered. 

[Helpful Tech Witch]

There's little difference between the likes of mx-4, nt-h1, and other high end compounds.

Technically, the "best" is a liquid metal solution.

[AlfonsLM]

If you really want to you can put the time in to flatten the surface so you can ring them together and have much better thermal properties than any thermal past could give you.

[freeagent]

I really like Thermalright TF8, TFX kicks butt too.. and it wont mess up your IHS and coldplate.

 

But I ran out and am now back to AS5 until I order more..

 

I am literally seeing roughly 3-4c difference at the very top end.

[Chris Pratt]

Assuming you get a good quality paste from the likes of Noctua, Artic, etc., they'll all perform about the same. One may be slightly better than another, but there's far more important factors that affect your temps more than what particular brand and model of high quality paste you use.

 

As for how much and application method, the answer is simply "enough". There's a ton of videos online comparing different methods like dot, cross, etc., and it virtually doesn't matter as long as you get good coverage. The easiest is just do a pea sized drop and call it a day. If you're new to this, there's nothing wrong with just doing some test applications. Apply the cooler, then take it off and see what the coverage was like. Clean up the die and adjust from there. Paste is relatively cheap and you get far more than you'll ever use in a tube unless you're building computers all day every day.

[RollinLower]

these guys did some seriously deep testing on the subject:

 

[Beerzerker]

1 hour ago, BillyGoat1776 said:

You may end up with several hundred different answers here.  I completely agree with @Skiiwee29 regarding margin of error.  Personally, I like the classic Arctic Silver 5 (cheap, available almost everywhere, works great).  Arctic MX-4 is carbon-based and non-conductive, works great, lasts a long time.  Kryonaut is excellent too (albeit a bit pricey).

As for the amount to apply, again I agree with @Skiiwee29... for Intel placing a dab in the middle of the CPU is good....for AMD, a bit more paste and, personally, I like to spread it with a guitar pick so that's it is nice and even across the entire surface of the CPU heat spreader.

The part I've bolded is true.
I too have ran many different TIM's along the way and each has it's own advantages and disadvantages. To know these things are important, based on the intended useage of the system with other factors thrown in too. Using a really cheap TIM is not a good idea, some of those just don't cut it and frankly, messy too.

Liquid metal right now is really popular inspite of the fact it's also "Corrosive" to the surfaces it makes contact with. This effect doesn't take too long to make itself known (2-3 years max) and if you're only expecting to have the pieces it's making contact with (CPU and Cooler) for about that length of time then it's fine if you don't mind it but if expecting long term service, it's a different story.

I've also seen many say "Kryonaut" and it does work, mostly without any issues but there have been a few instances (Some as of late) with it scratching CPU and cooler surfaces too.
Seems there have been bad batches of it at times that causes the scratching, however it's very unlikely you'll run into this. It's also more expensive vs other TIM's that have similar performance in a daily/gaming build and to me, that's not the ideal way to go if there is a cheaper alternative that works just as well under load within a margin of 2-3C at the most.
One more thing is Kryonaut was originally meant and formulated for sub-zero use which is why it's so expensive in the first place. Any TIM thats made to go sub-zero and give results is expensive per gram, no getting around it but there are cheaper alternatives even in that case which works about as well.

GeLid Extreme is a good TIM but beyond expensive for the amount you get and it too is prone to have bad batches appear a times, plus if it's been sitting on the shelf for too long it dries out, making it perform worse. A bad batch/container of it is "Dry" with a good container being goopy.

Other, well known TIM's such as MX-4, Noctua NT-H1 and others can and do just as well for any daily use for less cost.
Personally I've been running MX-4 Carbon and it's been a good all-around TIM for anything I've used it for and that includes sub-zero stuff. It's not "Thick" and hard to work with, spreads out quite well for the amount used per application and does give good results, plus it lasts a long time too. Not to mention with the WR I set recently, I was using MX-4 Carbon and that means it even beat any and all Kryonaut TIM'ed setups period or I woudn't have been able to do that in the first place.

NT-H1 is another that's good and not pricey either and it too lasts for a long time.

I don't know what the OP's expectations are for what they are doing but my point in all this is not just go with the hype, learn the differences between each to be considered and whatever works best in one's own particular case is what to go with.
 

[lerodemmy]

Really helpful answers.  Thanks!