The LAST Thermal Paste you'll ever need??

[IAnthonyFX]

49 minutes ago, AndPeroty said:

Did you apply thermal paste on vertical?

sometimes cooler clearance does not allow for normal application. In my case it requires a lot of maneuvering,

[RAM K]

11 hours ago, nicklmg said:

Buy IC Graphite on Amazon: http://geni.us/9pYB

 

You may NEVER need to use thermal paste again thanks to this new product...

 

 

Would it be possible to cut some of those pads so you can use it on a threadripper chipset?

[Sfekke]

I'll just stick to my tried and true Artic Silver 5, works mighty fine in all my laptops & pc's.

But for a test system I might actually buy a pack of these to see how they last under an extreme overclock for long periods; not in the near future though.

[ravenshrike]

1 hour ago, Zerim said:

Do you have a link to some numbers? 0.0009" doesn't seem unreasonable, and at that scale the Arctic Silver is already twice the resistance of the Dow TC-5022. I'm seeing numbers from Arctic Silver that the average is closer to 0.003"-0.005". 

Clearly now is the time for science. Somebody needs to turn on the Steve Signal and have him run prolonged 4 way tests between Kryonaut, AS5, DC 5022, and the IC thermal pad using a 1950X and a Vega 64.

 

https://www.moddiy.com/products/Dow-Corning-TC%2d5022-(3.5g).html

[trinitaet]

11 hours ago, nicklmg said:

11 hours ago, WereCatf said:

Quick question: did you guys happen to test if its orientation matters to its performance? If it conducts heat better in one direction than in an other, would it perform even a little bit better if it was orientated so that it conducts heat better towards the heatpipes? I'd love to know!

@AlexTheGreatish would be able to say for sure, but my understanding from the video is that it has different conduction ratings for horizontal & vertical conduction, rather than having different ratings for x & y. So that is to say I don't THINK it would matter, but I don't know for sure.

Since graphite basically is layers of hexagonal carbon atoms, there is no difference in "x" or "y". What matters is, if you measure the (thermal) conductivity along the layers or the conductivity perpendicular to them. The coupling between the atoms inside a layer is much stronger than the one between the layers and so, there will be a difference between the "z" axis and the ones, that span your layer.

[Spotty]

Once again LinusTechTips makes a clickbait title and fails to correctly identify an item as a pad. #thermalpastegate /sarcasm



 

[trinitaet]

I am really sorry to tell you this, but I am frankly a bit disappointed in you guys... Whenever there is some topic where you should do some scientific research up ahead, you almost never seem to do it. "...it's totally graphite. Uhm... which I guess is carbon..." just nailed the coffin there...

It is not as if you need to build you own physics lab or whatever, but it takes about 2 minutes of googling to find the properties of graphite, INCLUDING the baseline thermal conductivity. You do this for a living, please bother yourself with doing this!

[flexy123]

This is "interesting", albeit I love my MX4. (Even "normal" thermal pastes have come a long way since I built my first PC, where AS was all that we had, and it was conductive etc.).

 

BUT: I'd be highly interested how this graphite pad performs on GPUs and especially on delidded CPUs. (Of course, there is the issue of the pad being conductive). The reason I am asking...since pastes like MX4 which otherwise perform very good don't do as well, eg. for delidded CPUs.

 

Related question here, whether these pads are somehow reactive with other materials. Like Liquid Ultra which you are not supposed to use w/ pure copper HSFs.

 

That being said, interesting!

[JoostinOnline]

12 hours ago, Rasbir Singh said:

That's a broad claim. 

That's why the word "may" was used.

[WereCat]

4 hours ago, AndPeroty said:

Did you apply thermal paste on vertical?

Yes

14 hours ago, DrMacintosh said:

? You mount your coolers after they are in the case?

After motherboard is in the case? Yes. It is impossible to mount my cooler horizontaly because I have to screw it in from behind the motherboard.

[samcool55]

What about GPU's?

I mean most of them don't use an IHS and it's conductive, not a good combination...

[Tylerr]

2 hours ago, WereCat said:

Yes

After motherboard is in the case? Yes. It is impossible to mount my cooler horizontaly because I have to screw it in from behind the motherboard.

 

you can have your case laying on top of a table and slide off the side until you can reach the mounting parts.

[WereCat]

2 hours ago, Tylerr said:

 

you can have your case laying on top of a table and slide off the side until you can reach the mounting parts.

Or.. I can just mount it verticaly.

[Lady Fitzgerald]

I can't believe all the nit picking in this thread. So what if Linus referred to it as paste in the title of the video? It serves the same function. You nit pickers need to get over yourselves.

 

While I see possibilities for using it between the IHS and a cooler, tests I saw on another forum when using it between the die and the IHS were not as good. I really like the idea of not ever having to replace the TIM. For that, I would be happy to sacrifice a degree or two. I would prefer to see more tests done by others before committing to it.

 

Removing an installing a cooler while the MOBO is still inside a case ranges from just too much hassle to downright impossible due to the cramped quarters. For that reason, my next case will have a removable MOBO tray so I don't have to wrestle with the entire case.

[Jtalk4456]

54 minutes ago, Lady Fitzgerald said:

I can't believe all the nit picking in this thread. So what if Linus referred to it as paste in the title of the video? It serves the same function. You nit pickers need to get over yourselves.

 

While I see possibilities for using it between the IHS and a cooler, tests I saw on another forum when using it between the die and the IHS were not as good. I really like the idea of not ever having to replace the TIM. For that, I would be happy to sacrifice a degree or two. I would prefer to see more tests done by others before committing to it.

Agreed. 

To respond to some questions about how well it would deal with threadripper, I'd love to see these used on a threadripper in a different video and maybe they have to "glue" the edges together with traditional paste, not sure. In either case, I would love for them to make a permanent build with these pads and use for about a year and see how it held up long term

[AlexTheGreatish]

21 hours ago, Hunter259 said:

Is it able to replace traditional thermal pads? Does it have higher conductivity than a normal thermal pad?

I wouldn't, thermal pads are generally a lot thicker so you might not get proper contact.

20 hours ago, Enderman said:

Are you guys going to switch all your testing to these reusable thermal pads because they perform better?

Or stick to thermal paste because that's what 99.9% of people are actually using?

 

Also yeah, as other people have mentioned graphite thermal pads have existed for a few decades, I'm surprised it actually worked better than thermal paste because usually they are crap on anything that isn't perfectly flat. Nice to see it actually work.

 

Still worse than liquid metal though

Probably not, we have heaps of thermal paste which isn't conductive and you really need to screw up the application to have a measurable difference in temperature.

10 hours ago, RAM K said:

Would it be possible to cut some of those pads so you can use it on a threadripper chipset?

Yup

[DoctorWho1975]

Video sponsored by lmgtfy.com

[DarkSwordsman]

I wonder if the thermal conductivity has to do with thickness? That's mainly because of how wide the X and Y are and how thin the Z is. Alex mentioned that the thermal conductivity X and Y is far greater than Z.

It's a long shot, but what if you stack 5, 10, or even more of them? Would it increase the thermal conductivity of Z?

[yenic]

First conductive product that I'll be using regularly on my machines. As far as microscopic pits not being filled by this, people get too worked up about that idea. It's either not a problem as their results show or possibly even non-measurable, as their 1C difference is probably due to the material used.

 

Another MX4 fan here. I use it on everything including die to IHS with no issues. In my tests with non-conductive pastes, MX4 is about as good as it gets unless you're going below ambient temp. This product could be the first thing to move me off MX4, especially since I no longer have to delid every CPU I buy out of the box, thanks to Ryzen.

[Sacredsock]

I'd like to see a follow up video to this, having skimmed through the three pages so far there does seem to be some interesting points raised.

 

Personally I'd be interested to see how it holds up to being used in a standard system, ie vertically positioned between CPU and heatsink. 

The other thing that intrigues me is could it be used in delidding? maybe sacrifice a slightly older 115x chip for testing?

 

[Deimos F]

Hey Alex, here's a crazy idea: using a combination of this pad AND some thermal paste, for those really wide processors. Hear me out:

 

Big enterprise processors have big lids with considerable area. I expect certain spots along the lid will be hotter than others, depending on what part of the chip happens to be below. A cooler tries to "remove" the heat as fast as it can, but the amount of heat being transferred across a specific point has a limit. Now, since some areas of the lid have more heat to lose than others, and since that pad has a huge XY conductivity, WHAT IF you applied a tiny amount of thermal paste on the lid (half the usual amount), then slapped an adequately sized pad on top, then added the other half of the thermal paste on top, and then placed the cooler?

My reasoning for this is the following: according to the test results, the Z axis conductivity of the pad has similar performance to decent thermal paste, though whether this is due to imperfections in surface contact isn't made clear). However the XY conductivity is much much higher, and that characteristic isn't seeing much use. So in regards to the Z axis, there would be no difference in conductivity, as long as the amount of total material present doesn't increase significantly (which it shouldn't due to the mount pressure). However, any hot spots with a higher amount of heat would have that heat very rapidly distributed across the entire area of the lid, theoretically increasing the efficiency at which heat is passed from the lid to the cooler.

 

I mean, it sounds a bit extreme, but isn't that the whole point of channels like Linus Tech Tips? To spend the time and effort into testing crazy stuff that no ordinary citizen would bother with?

[mariushm]

55 minutes ago, Deimos F said:

...

On Threadripper and EPYC processors they use solder to attach the top lid to the silicon chips inside, so heat transfers very well from the chips to the top lid.

On Threadriper, the working dies are arranged diagonally, two are working and two are fake and are there just to prevent the top lid from flexing downwards, to stay as flat as possible.

So already heat is transferred on the x-y axis quite well, since there's a very good physical connection between dies and the top. Now you only want to transfer the heat on the z axis as efficiently as possible and this is where you have a problem with large area chips ... the tops can still be either concave or convex (the surfaces are not perfectly flat). So having an abundance of thermal paste between the cpu top will help 

See https://www.gamersnexus.net/news-pc/3008-threadripper-cooler-and-thermalpaste-coverage-vs-die-ihs

 

On the z axis, these thermal pads are not great, some pastes have better thermal conductivity and the ability of pastes to be squished and spread really helps.

[maxichrome]

23 hours ago, AndPeroty said:

The problem with a delidded CPU may be that ir is conductive.

True, Cut down to size, it shouldn't be a problem (like small application of liquid metal)

[tikker]

1 hour ago, VegetableStu said:

is that a one-time use thing? o_o

Well it melts, so... yeah

[Cromartie]

On 4/29/2018 at 11:48 AM, nicklmg said:

Buy IC Graphite on Amazon: http://geni.us/9pYB

 

You may NEVER need to use thermal paste again thanks to this new product...

 

 

Hey Linus! You know how you said you were having doubts for the size of the pads, say for a threadripper? well check this out:

https://www.amazon.com/gp/offer-listing/B07CK9SHZG/ref=olp_twister_child?ie=UTF8&colid=3SC0I57G6IMTX&coliid=I3RYJDVC9Y0JNM&mv_size_name=1

They offer it in 40x40mm too!