Alphacool Unveils Apex 17 W/mK Thermal Paste

[btarunr]

Alphacool introduces the new Apex thermal paste. With a thermal conductivity of 17 W/mK, it enters the circle of the most powerful thermal pastes on the market. The main objective during development was, of course, to increase the thermal conductivity. However, points such as viscosity and durability should not be ignored. Alphacool was able to optimize the viscosity so that the Apex thermal paste is particularly suitable for high contact pressures and is still easy to process. The main ingredient of Apex Thermal Grease is a nano-powder with a very low thermal impedance. This makes it possible for electronic components to be cooled efficiently, thus extending their durability and improving their reliability.



The thermal paste is delivered in a 4 g applicator with a screwable cap. The supplied amount is sufficient for several applications and can be stored in the applicator to protect it from drying out. The Apex 17 W/mK Thermal Grease will be available in the Alphacool online store (item number 1022240) from 06/16/2022.



View at TechPowerUp Main Site

 

[Yttersta]

Here is a realistic depiction of events which took place in Alphacool Marketing department a while ago:

Product guy: "Guys! Here's the new product; Apex. Our best thermal paste to date, it is to compete with the likes of Arctic, Noctua, Thermal Grizzly and the lot at the very top end. Make sure you take some pictures of it for the PR"
Marketing guy: "Sure thing! Let us bring out some appropriate hardware to go along our high performance paste. These should be representative of current day components as well of course!"
...
Marketing guy slaps the top of 11 year old motherboard box: Well, these must!


I mean, not that it matters much for a thermal paste but SB-e, IB-e weren't that difficult to cool with far less component density and lower TDPs than parts of today, with huge heat spreaders and low clocks (compared with today's average top tier).

 

[Crackong]

Wow that is a Sabertooth X79 MB..what a relic

 

[dgianstefani]

Wonder how this will fare against Kryonaut Extreme.

 

[EddyAlphacool]

@Yttersta
As global Head of Marketing, I'll tell you this.

Our test team has stolen all our equipment so they can do extensive testing on the new and upcoming products. Because there are currently a lot of new products in the works that will help with the conversion and realignment of our brand.
So I had a choice. To take an old board, just to be shown how to apply it, or to spend money on a new motherboard. Of course, such expenses have to be added back into the end customer prices when they add up.
With the picture we do not want to advertise expensive high-end boards, but show how our product looks and can and should be used. Nothing else.

Besides..... being a bit retro is also modern today.

 

[DeeJay1001]

Wow that is a Sabertooth X79 MB..what a relic

One of the great ones. I bought one second hand with a xeon cpu that I still have running as a media server. 5 years of basically 24/7 service. Never skipped a beat.

 

[Valantar]

12W/mK sounds impressive - though sadly AFAIK there's no accepted standard for measuring this, so W/mK numbers across manufacturers are essentially nonsense. Still, Alphacool makes a lot of good stuff, so I'd be interested in seeing how this performs.

 

[EddyAlphacool]

@Valantar
The W/mk value is a standard value calculated on the basis of the substances and materials used. Nothing needs to be measured. As an example, the thermal conductivity of copper is always identical with the same R-unit value.

 

[ir_cow]

Anyone notice the X79 MB?

 

[Yttersta]

@Yttersta
As global Head of Marketing, I'll tell you this.

Our test team has stolen all our equipment so they can do extensive testing on the new and upcoming products. Because there are currently a lot of new products in the works that will help with the conversion and realignment of our brand.
So I had a choice. To take an old board, just to be shown how to apply it, or to spend money on a new motherboard. Of course, such expenses have to be added back into the end customer prices when they add up.
With the picture we do not want to advertise expensive high-end boards, but show how our product looks and can and should be used. Nothing else.

Besides..... being a bit retro is also modern today.

Your only mistake was not to take a picture of yourself slapping the top of the box for the memes!

We want pics or else! :-D

Jokes aside, loved the X79 days. They were simpler days, wouldn't you agree? We just had the 40% YoY perf jump of Sandy Bridge, then got SB-e, which had even more memory bandwidth than the top Nehalems, and the 6c/12t 3930k was very well priced! They cooled easily, overclocked well, endured for years and years. In fact, the longest serving system I've ever had was on X79, it was a beast. From 2012 till 2018, I used it.

 

[Valantar]

@Valantar
The W/mk value is a standard value calculated on the basis of the substances and materials used. Nothing needs to be measured. As an example, the thermal conductivity of copper is always identical with the same R-unit value.

Sure, except that copper is a solid and has mostly the same characteristics across a huge range of environmental conditions, while thermal pastes are complex mixed fluids with characteristics that change depending on the implementation and inherent characteristics of the paste (particle size, viscosity, cure time, etc.) that make even more of a mess of this. You can absolutely measure a neutral W/mK reading for a paste by measuring heat flux through a given thickness of paste, but that number doesn't necessarily transfer easily to a CPU/GPU cooling application. Heck, your own marketing says as much, talking about the balance of viscosity and thermal transfer and this being optimized for high mounting pressures.

For a functional W/mK reading for computer cooling applications, this number needs to be measured in a computer cooling application. And there is no standardized way of doing so - which is why you get pastes that perform the same despite drastically different W/mK ratings. Still, your rating is on the high side for a paste, and your explicit mention of spreadability makes me hopeful that it's not too thick - though "optimized for high mounting pressures" tends to mean it's not the most spreadable.Still, as I said: looking forward to seeing this tested!

 

[Sabishii Hito]

@Valantar
The W/mk value is a standard value calculated on the basis of the substances and materials used. Nothing needs to be measured. As an example, the thermal conductivity of copper is always identical with the same R-unit value.

Hope it works better than this stuff.

 

[FreedomEclipse]

@EddyAlphacool how sure can we be sure that the thermalpaste wont cause any pitting/scratching to mating surfaces like two other manufactures that have been through the hands of TPU members?

Cuz i see the word 'nano-particles' and i start to get worried. The pastes that did the damage were made up of nano particles too - albeit diamond ones.


If the worst case scenario does happen and the paste does damage something, Will Alphacool compensate people?

 

[firewrath9]

@EddyAlphacool how sure can we be sure that the thermalpaste wont cause any pitting/scratching to mating surfaces like two other manufactures that have been through the hands of TPU members?

Cuz i see the word 'nano-particles' and i start to get worried. The pastes that did the damage were made up of nano particles too - albeit diamond ones.


If the worst case scenario does happen and the paste does damage something, Will Alphacool compensate people?

Have CPUs stopped working because of pitting? Usually they're like really small and literally dont affect anything. Anyone got data on how this affected performance/temps/etc.?

 

[FreedomEclipse]

Have CPUs stopped working because of pitting? Usually they're like really small and literally dont affect anything. Anyone got data on how this affected performance/temps/etc.?

No but damage is damage, Not to mention if you were to use the paste on a bare die.

If you take your car to you local detailer and they scratch up all the paint work and scuff the interior - does it make it ok because it doesnt affect anything?

 

[Dr. Dro]

@Yttersta
As global Head of Marketing, I'll tell you this.

Our test team has stolen all our equipment so they can do extensive testing on the new and upcoming products. Because there are currently a lot of new products in the works that will help with the conversion and realignment of our brand.
So I had a choice. To take an old board, just to be shown how to apply it, or to spend money on a new motherboard. Of course, such expenses have to be added back into the end customer prices when they add up.
With the picture we do not want to advertise expensive high-end boards, but show how our product looks and can and should be used. Nothing else.

Besides..... being a bit retro is also modern today.

Only we'd know how much these older Intel HEDT chips can heat up, so i'll chalk the motherboard choice up as a W, my good man.

I'd love to see these on the Brazilian market, in fact, i'd love to see all of your company's products available here someday. It gets... "warm" here during summer, after all.

Have CPUs stopped working because of pitting? Usually they're like really small and literally dont affect anything. Anyone got data on how this affected performance/temps/etc.?

Mechanically and functionally, you may even be fine, the problem is that a dented surface may form small pockets of air that will function as a thermal isolator, severely degrading cooling performance. If that becomes severe enough, you'll end up needing to lap both the IHS and the cold plate to ensure a smooth contact.

 

[EddyAlphacool]

@Yttersta

I am a secret person who does not show myself in front of cameras. At least I try to avoid that with all my might. Untypical for a marketing person, I know, but I always have the feeling that when PR people put themselves too much in the foreground they do their job badly.
Besides..... I would just scare people.

@Valantar
You are right about almost everything, but the pure W/mk value is a calculation that is independent of the viscosity. The W/mk value is a purely technical value. How the whole thing then behaves on a GPU or CPU is of course dependent on various factors that you have listed.
Otherwise, a 17W/mk thermal pad would be just as good as a corresponding thermal paste, which is not the case as we know.

@Sabishii Hito
The one in your picture is an industrial sponsor. So for mass applications. You can't compare it with the Apex.

@FreedomEclipse
Honestly, I don't understand the problem. If you use liquid metal, for example, the surface of the CPU will also be attacked.
I'm not aware of anything like that happening in our tests.

 

[Valantar]

@Valantar
You are right about almost everything, but the pure W/mk value is a calculation that is independent of the viscosity. The W/mk value is a purely technical value. How the whole thing then behaves on a GPU or CPU is of course dependent on various factors that you have listed.
Otherwise, a 17W/mk thermal pad would be just as good as a corresponding thermal paste, which is not the case as we know.

That is true, but that also illustrates why the baseline W/mK number is borderline useless for comparing thermal pastes - and why some manufacturers choose to calcuate theirs in situ rather than as an implementation-independent measurement. It's one number out of several variables that are always present and highly variable - and, as I started out saying, there's no agreement on how to compare this. If everyone agreed on providing the base W/mK measurement that would have some use, but it would of course still be dependent on the physical characteristics of the paste. Still looking forward to seeing this reviewed though - it's been a while since I've seen a new attempt at a high thermal conductivity paste, with most recent ones (MX-5, NT-H2, etc.) all aiming for spreadability over absolute conductivity.

 

[FreedomEclipse]

Honestly, I don't understand the problem. If you use liquid metal, for example, the surface of the CPU will also be attacked.
I'm not aware of anything like that happening in our tests.

discolouration of the mating surfaces isnt the same as scratching/pitting.

Thermal Grizzly thread here

IC Diamond thread here

Hats off to TG - they sorted the whole issue out without any pushback. There were multiple reports of pitting/scratching with their pastes on TPU and TG sorted it all out real quick.

The guy from IC Diamond however (who i think was also the CEO) accused multiple TPU members and another TPU member who sent their IC Diamond sample off to an independent lab for analysis of DOCTORING the images. Lots of drama and denial and insults thrown around at the community by that man.


-- In the case of Liquid Metal, if you short out your own hardware because you didnt take precautions to protect the surrounding hardware should the liquid metal leak or drip out then thats your own mistake. Even if the liquid metal does start some sort of corrosion within the IHS or the heatsink, it will take a long long time before any of those will be beyond use. I have so far only heard of discolouration though as the Liquid metal gets slowly absorbed by the copper used for the heatsink or IHS.



Simple question - If we use your product and it does something like this:

What will Alphacool do?

The damage you see here was because the 'nano-particles' wasnt 'nano' enough and when you mount the the heatsink onto the CPU/GPU and screw everything all down. The larger particles get pressed against the mating surfaces which cause damage.

This is only one example though - There are more in the threads posted.... Just dont be that Guy from IC Diamond and make sure that Alphacool QC every single batch before boxing and shipping to suppliers.

 

[EddyAlphacool]

@Valantar
As I said, I understand what you mean. But the W/mk value is simply a technical and generally valid specification. How it is calculated is standardized. Of course, you would like to have a "practical" value. But we'd rather leave that to independent testers.
Otherwise, you'll have a similar disaster as the noise measurements of the fires for fans. I can tell you how ours are determined, but in the end you can't compare it with other manufacturers who determine this value differently. Here, for example, there was once a manufacturer who did it quite cleverly with the dBA data. They have the difference between the background noise of their test lab (which is always present) and the actual volume indicated. Then there were suddenly values of 11 dBA. But a normal room (not an anechoic chamber) that is supposedly dead quiet already has a background noise of ~ 24-26dBA.

@FreedomEclipse
As I said, this did not even happen during our tests. At least not that I am aware of. Therefore, we did not have to think about how to handle such cases.

 

[ThrashZone]

Hi,
Any dates for stock on performance pc/ newegg or amazon yet ?

 

[EddyAlphacool]

@ThrashZone
Sorry, I can't tell you anything about that. I don't have anything to do with sales directly. But I am sure that our sales department is working hard so that the product will soon be available everywhere.

 

[Valantar]

@Valantar
As I said, I understand what you mean. But the W/mk value is simply a technical and generally valid specification. How it is calculated is standardized. Of course, you would like to have a "practical" value. But we'd rather leave that to independent testers.
Otherwise, you'll have a similar disaster as the noise measurements of the fires for fans. I can tell you how ours are determined, but in the end you can't compare it with other manufacturers who determine this value differently. Here, for example, there was once a manufacturer who did it quite cleverly with the dBA data. They have the difference between the background noise of their test lab (which is always present) and the actual volume indicated. Then there were suddenly values of 11 dBA. But a normal room (not an anechoic chamber) that is supposedly dead quiet already has a background noise of ~ 24-26dBA.

Yep. Which is exactly why there ought to be standardized testing methods for these things, with clearly defined testing parameters.

 

[howiec]

Can anyone confirm that the Apex paste is not electrically conductive?

I'm considering using this for future GPU paste replacement if tests show good performance and maintains it over time (doesn't dry out too quickly).

Thanks!

 

[Dr. Dro]

Simple question - If we use your product and it does something like this:

What will Alphacool do?

The damage you see here was because the 'nano-particles' wasnt 'nano' enough and when you mount the the heatsink onto the CPU/GPU and screw everything all down. The larger particles get pressed against the mating surfaces which cause damage.

This is only one example though - There are more in the threads posted.... Just dont be that Guy from IC Diamond and make sure that Alphacool QC every single batch before boxing and shipping to suppliers.

Yeah, that is pretty nasty, oof. It would be awesome if it's ensured such "nano-particles" of ultra-hard compound didn't make our chips look like they went through the meatgrinder

And if I may, reading that IC Diamond thread was some high quality entertainment. What was the guy smoking? No wonder they haven't been seen for 9 years now. Must still be ashamed of that one. I must confess I love a big ol' forum fight!