News Thermaltake Wants to Jazz up the Way You Apply Thermal Compound

Giroro

Honorable
Jan 22, 2015
687
127
11,190
13
A stencil makes a lot of sense.
It makes so much sense that most of their competitors' R&D have probably already tried/rejected this idea, for some reason.
 

Phaaze88

Glorious
Ambassador
WTH is this garbage?
That is such a waste of paste... this has me wanting to pull my hair out.
Screw you, Thermalfake.

As long as the paste covers the area of the die underneath, that's all that matters. Why all this?
 
Reactions: madmatt30

madmatt30

Titan
Ambassador
WTH is this garbage?
That is such a waste of paste... this has me wanting to pull my hair out.
Screw you, Thermalfake.

As long as the paste covers the area of the die underneath, that's all that matters. Why all this?
Its the kind of tat that would appeal to a complete novice maybe?

Problem being that kind of novice would probably buy a prebuild rendering the whole idea pointless.
 

Phaaze88

Glorious
Ambassador
Its the kind of tat that would appeal to a complete novice maybe?
Why, of course, take advantage of people who wouldn't know any better... how typical.
/cynicism

It makes sense to me, the stencil's shape can ensure complete coverage...
That's the thing. Complete coverage of the IHS is pointless. The user only needs to apply enough - bar cooler mounting pressure - to cover the silicon dies beneath the IHS.
This product is expensive, as it serves to make users spend more paste than needed per serving.
None of today's cpus have dies that take up the entirety of the IHS they're under. If the user has a general idea, or actually knows the size and location of the die(s), they can avoid wasting paste needlessly.
This junk here goes in the opposite direction.

Intel's monolithic dies? The center application will take care of them. For the larger X series dies, a short line horizontally or vertically(depends on the cooler) and they'll be good.
AMD's Ryzen 3000? 3 dots will cover a 3950X. 3600? 2 dots - one at the side, and another in a corner.
Threadripper? Do a big X or +.
 

PapaCrazy

Distinguished
Dec 28, 2011
223
33
18,740
7
That's the thing. Complete coverage of the IHS is pointless. The user only needs to apply enough - bar cooler mounting pressure - to cover the silicon dies beneath the IHS.
This product is expensive, as it serves to make users spend more paste than needed per serving.
None of today's cpus have dies that take up the entirety of the IHS they're under. If the user has a general idea, or actually knows the size and location of the die(s), they can avoid wasting paste needlessly.
This junk here goes in the opposite direction.

Intel's monolithic dies? The center application will take care of them. For the larger X series dies, a short line horizontally or vertically(depends on the cooler) and they'll be good.
AMD's Ryzen 3000? 3 dots will cover a 3950X. 3600? 2 dots - one at the side, and another in a corner.
Threadripper? Do a big X or +.
I used to think this until I heard Debaur say it was more effective to cover the entire IHS by spreading out compound so the whole surface could conduct heat. Not what I expected to hear, but I think he has more experience than anyone.
 

Phaaze88

Glorious
Ambassador
I used to think this until I heard Debaur say it was more effective to cover the entire IHS by spreading out compound so the whole surface could conduct heat. Not what I expected to hear, but I think he has more experience than anyone.
Well, ok. Do what works for you.

I mean, it's not like I didn't personally test different paste applications myself to sate my own curiousity... how hard is that?
I could just go and take some popular(?) individual's... opinion(?) - IDK what to call it - as fact and not bother, right?
 
Last edited:

AnimeMania

Honorable
Dec 8, 2014
317
15
10,815
13
Unless you have a job building computers, most people who build a computer will be building just the one. It is better to do it right the first time, cover the entire CPU with the right amount of thermal paste. Once the cooler is properly seated and working, everything left in the tube is pretty much wasted paste. I have a really old computer and had to repaste the CPU once in 8 years. I used the paste that originally came with the computer, opened and not used for 6 years, it seemed fine.
 

madmatt30

Titan
Ambassador
I used to think this until I heard Debaur say it was more effective to cover the entire IHS by spreading out compound so the whole surface could conduct heat. Not what I expected to hear, but I think he has more experience than anyone.
It depends on the paste used imo, arctic silver and mx are fairly runny, a few dots of those and it will spread out fairly evenly.

The particle based pastes like kryonaut grizzly and cm mastergel maker come with spreading spatulas for a reason, theyre too dense to spread just with heatsink pressure.

I absolutely spread virtually the entire heatsink when using these pastes, why wouldn't you take advantage of that huge soldered heatspreader on the ryzen chips?

This obviously is dependant on the heatsink actually being used, some older universal heatsinks dont come close to covering the ryzen heatspreaders in full, they're big chips compared to older intel and there can be a fair portion of heatspreader left showing .


The thermaltake paste is probably a similar consistency, its just the stencil I see no practical use for personally.

I also dont get who that paste and system is aimed at, stock coolers come with paste preapplied, aftermarkets come with it preapplied or included.

If you're buying it to repaste a stock cooler (especially on a ryzen) its an incredibly flawed system because the spire and stealth coolers don't come close to covering the cpu heatspreader meaning youll have thermal compond just sitting loose and exposed which isnt a great idea.

Anybody buying that thermaltake paste seperately is buying it for a reason but Im not sure what that reason would be.
 

watzupken

Proper
Mar 16, 2020
176
45
110
0
WTH is this garbage?
That is such a waste of paste... this has me wanting to pull my hair out.
Screw you, Thermalfake.

As long as the paste covers the area of the die underneath, that's all that matters. Why all this?
Totally agree with what you mentioned. In this case, I feel Thermaltake has gone too far trying to differentiate their products from others.
 

Phaaze88

Glorious
Ambassador
Nothing beats testing it yourself if you're curious enough - screw what others say how it should be done, yeah?
I've been using NT-H1 and NT-H2 - the former spreads more easily - and don't notice any difference between a full spread and 'just enough to cover the die beneath'.
I'll see a more noticeable difference from removing one of my NH-D15S' fans...
 
WTH is this garbage?
That is such a waste of paste... this has me wanting to pull my hair out.
Screw you, Thermalfake.

As long as the paste covers the area of the die underneath, that's all that matters. Why all this?
Even a broken clock is right twice a day.

I don't think it a horrible idea. In fact, I like it. A consistent thickness of dots ensures a more even application of paste. When you hand spread with the credit card method, you risk having one side thicker than another. Consistent coverage should help with complete coverage. If one side is too thick, there's a risk of one side having greater mounting pressure then another and the paste not spread out evenly for complete coverage.

But their photos of finished spread really shows an uneven application on the AMD side. Look at the corners. They are bare. I think this is more a problem of the IHS/Cold plate though. Corsair's H60 was complete garbage in terms of a cold plate being flat. I got two in a row that were so uneven that complete coverage was near impossible without an extra tight bolt down which lead to a snapped screw head once.
 
Last edited:
Well, ok. Do what works for you.

I mean, it's not like I didn't personally test different paste applications myself to sate my own curiousity... how hard is that?
I could just go and take some popular(?) individual's... opinion(?) - IDK what to call it - as fact and not bother, right?
I'm an aerospace engineer (and coder), and have extensive thermodynamics background. Currently one of my main task is studying heat transfer on coils. Is uses a lot of complex finite element differential math analysis to simulate heat transfer using NIST and industry standards. (I have other skills and duties, but only listing the one that is relevant.)

Complete coverage does help but exponentially decays the further from the heat source. Good thermal conductivity is pretty high up there. Trapped Stagnant Air has a rotten thermal conductivity (And this why we have multiple panes of glass windows.) So it's essential you do get complete thermal coverage. HOWEVER, paste thickness really doesn't have that great an impact on the heat transfer model because the layers are so relatively thin. This last statement is dependent that you do get a full coverage spread however.

Gamers Nexus did a great demonstration where they left a wafer thin layer and a really thick one and the difference came out ~ 1 C in their testing IIRC.
 

Phaaze88

Glorious
Ambassador
I don't think it a horrible idea.
It is a horrible idea though, because the one con outweighs the benefits of this product: Expensive though paste waste.
Thermal pads would be more cost effective and more efficient - except for high TDP cpus - and less wasteful!

But their photos of finished spread really shows an uneven application on the AMD side. Look at the corners. They are bare. I think this is more a problem of the IHS/Cold plate though.
...
Look how ridiculously THICK those applications are. We, as users, are just trying to fill in some pits that can't be seen with the naked eye.
Too. Much. Paste.
It should look more like THIS after mounting a cooler.

I'm an aerospace engineer (and coder), and have extensive thermodynamics background. Currently one of my main task is studying heat transfer on coils. Is uses a lot of complex finite element differential math analysis to simulate heat transfer using NIST and industry standards. (I have other skills and duties, but only listing the one that is relevant.)

Complete coverage does help but exponentially decays the further from the heat source. Good thermal conductivity is pretty high up there. Trapped Stagnant Air has a rotten thermal conductivity (And this why we have multiple panes of glass windows.) So it's essential you do get complete thermal coverage. HOWEVER, paste thickness really doesn't have that great an impact on the heat transfer model because the layers are so relatively thin. This last statement is dependent that you do get a full coverage spread however.

Gamers Nexus did a great demonstration where they left a wafer thin layer and a really thick one and the difference came out ~ 1 C in their testing IIRC.
That's great!

Thin layer, thick layer... margin of error difference... I do remember bits and pieces of the video you're referring to.
Now, where's the sample thermal difference for my argument of 'just enough to cover the die(s) beneath'? You're not going to leave that out, are you?
 

gg83

Honorable
Jul 10, 2015
157
21
10,585
0
Has anyone noticed how particular people are about thermal compound applications? I think Steve @ GN likes to squeeze the whole tube onto one cpu just to laugh at the reactions
 
It is a horrible idea though, because the one con outweighs the benefits of this product: Expensive though paste waste.
Thermal pads would be more cost effective and more efficient - except for high TDP cpus - and less wasteful!


...
Look how ridiculously THICK those applications are. We, as users, are just trying to fill in some pits that can't be seen with the naked eye.
Too. Much. Paste.
It should look more like THIS after mounting a cooler.


That's great!

Thin layer, thick layer... margin of error difference... I do remember bits and pieces of the video you're referring to.
Now, where's the sample thermal difference for my argument of 'just enough to cover the die(s) beneath'? You're not going to leave that out, are you?
By no means. A thin application is a more thrifty use of the paste. Just bear in mind if you get too thin there is a risk of lack of proper compelete coverage due to uneven surfaces. I rank consistent even and complete coverage more critical than thickness.

If I buy a quality paste i get two or three solid applications per tube. So thats about $5 per application. When im building a $500+ system that's trivial cost addition in my book.
 

Phaaze88

Glorious
Ambassador
By no means. A thin application is a more thrifty use of the paste. Just bear in mind if you get too thin there is a risk of lack of proper compelete coverage due to uneven temps.
Sure, within 10C of the hottest and coolest cores - at least, according to Intel. IDK what AMD considers normal though.

If I buy a quality paste i get two or three solid applications per tube. So thats about $5 per application. When im building a $500+ system that's trivial cost addition in my book.
That's going to depend on the individual then. 2-3 applications is too few for my use case, that's why I buy 10g of NT-H2 these days.
This Thermaltake product does not target me, as it does nothing for me:
-more expensive overall VS what I already use
-the 'guide'(stencil) VS me having no problem experimenting different applications with the paste I already use
 
Sure, within 10C of the hottest and coolest cores - at least, according to Intel. IDK what AMD considers normal though.


That's going to depend on the individual then. 2-3 applications is too few for my use case, that's why I buy 10g of NT-H2 these days.
This Thermaltake product does not target me, as it does nothing for me:
-more expensive overall VS what I already use
-the 'guide'(stencil) VS me having no problem experimenting different applications with the paste I already use
You're a busy guy Phase to use that much paste. I think I only put together 20 some systems over my lifetime. But I do respect your opinion.

BTW: I made a typo in my previous post. I meant to say uneven Surfaces. Not uneven temps.
 

ASK THE COMMUNITY

TRENDING THREADS