ASUS brings "Exotic Thermal Compound" to their Gaming Notebooks

[WYP]

Up to 13 degrees cooler than normal thermal paste.



Read more about ASUS' use of liquid metal thermal compounds on some of their latest ROG notebooks.

 

[NeverBackDown]

Wow that's impressive. Laptops obviously don't have as much surface area for cooling and it's still able to drop 13C. That's maximizing efficiency. Pretty cool they solved the spillage problem to(theoretically)

 

[Warchild]

13C is an impressive drop in any scenario. For laptops its almost too good to be true. Heres hoping for mainstream gaming laptops to follow suit

 

[dazbobaby]

I've often thought of doing the same with my laptop, both the CPU and GPU can be got at so that's not the problem.


The biggest issue is the possibility/likelyhood that it will spill out over time, because you just can't keep a laptop flat all the time. Electrical tape or even PTFE will help... but it's still a risk.


Nice to see ASUS finally cracked that problem.

 

[Warchild]

I've often thought of doing the same with my laptop, both the CPU and GPU can be got at so that's not the problem.


The biggest issue is the possibility/likelyhood that it will spill out over time, because you just can't keep a laptop flat all the time. Electrical tape or even PTFE will help... but it's still a risk.


Nice to see ASUS finally cracked that problem.

Dont use tape. With heat, tape can still warp. Stickiness can melt and slide. If you know someone or have got youself some clear nail varnish. Use that. Let it cure, and its rock solid for years to come.

Drawback is that there is no going back once done. (Not that it matters)

 

[Avet]

The biggest issue is the possibility/likelyhood that it will spill out over time, because you just can't keep a laptop flat all the time. Electrical tape or even PTFE will help... but it's still a risk.

Liquid metal TIM has very high surface tension. That is why it is such a pain to spread out, and why you need to apply it on both surfaces. It will never leak out if you apply a proper amount. If it does leak out it means you have put way to much.

And also pretty much everyone is applying it wrong, so people are afraid of using it.

This is how you should do it.

https://www.youtube.com/watch?v=Vp-NY5x7ymg

 

[Warchild]

And also pretty much everyone is applying it wrong, so people are afraid of using it.



https://www.youtube.com/watch?v=Vp-NY5x7ymg

Err disagree there. Alot may do it wrong, but not everyone. Even the guy in the video used too much. All you need is a ridiculously thin layer to fill in the imperfections of either surface

 

[Avet]

Err disagree there. Alot may do it wrong, but not everyone. Even the guy in the video used too much. All you need is a ridiculously thin layer to fill in the imperfections of either surface

The amount can vary from CPU to CPU. I applied to little on my first try, and then added just a smidge more, and it was perfect. It is maybe a tiny bit too much in this video, but there is no pooling, which is important. Also IHS wasn't lapped, and they can be quite uneven. Die itself can be concave, or convex. If you don't sand down the bit of IHS that usually has silicon adhesive there is still a gap between IHS and the die. If you would do direct die, or custom IHS, lapped IHS, then yes this would be too much.

The main point of the video was to apply LM to the cue tip, and not try to squeeze it on CPU.

 

[trawetSluaP]

Uh oh, I only applied LM to my CPU and not the back side of the IHS...

 

[dazbobaby]

Thanks Warchild and Avet.


Good advice from you both

 

[Thanos]

What power are we talking here for a 13C improvement?

What is a standard TIM?

It says 13C is a theoretical value, does that mean at 200W CPU theoretical power draw?

Would be nice to see a direct comparison between the likes of say Gelid Extreme at the same power levels.

Instead of improving the cooling system of their G703GXR for Intel's 9th Generation processors, ASUS has increased the conductivity of the interface between their notebook cooler and the system's processor, allowing heat to be more efficiently transferred to the system, making the user's CPU cooler, and the entire system quieter.

This seems to be suggesting they are using the same cooler for both 6 core 8950HK and 8 core 9980HK. Isn't the 8 core going to draw more power, is just the TIM really that good to not only cancel the effect of running about 30% more power but to reduce fan speed and still run the CPU cooler to boot?

 

[Avet]

What power are we talking here for a 13C improvement?

10-20C improvement is what you would expect depending on CPU and cooling solution. Whatever the number it is more than enough to prevent thermal throttling.

What is a standard TIM?

Standard TIM is probably something really cheap. Basically any paste that you replace it with will do better.

It says 13C is a theoretical value, does that mean at 200W CPU theoretical power draw?

They say 13C theoretical value, because results may vary from model to model, and CPU to CPU. So if they say it is 13C better, and someone does not get it there would be complaints. They are covering their behinds.

Would be nice to see a direct comparison between the likes of say Gelid Extreme at the same power levels.

Try using google sometimes. https://www.tomshardware.com/reviews/thermal-paste-comparison,5108.html

This seems to be suggesting they are using the same cooler for both 6 core 8950HK and 8 core 9980HK. Isn't the 8 core going to draw more power, is just the TIM really that good to not only cancel the effect of running about 30% more power but to reduce fan speed and still run the CPU cooler to boot?

There is not much choice when designing a laptop. You have very little space, and they try to design the best cooler that can fit. To reduce cost it is used on everything. And also there are no silent gaming/high-end laptops. If you put anything that even resembles power it will scream. Not even the mighty Alienware Area 51 can be silent.

 

[looz]

What makes standard TIM worse than enthusiast products isn't necessarily them skipping on the cost, but stability. It's average out of the factory and it's going to be average 5 years from now. An enthusiast paste might dry up.

 

[Avet]

What makes standard TIM worse than enthusiast products isn't necessarily them skipping on the cost, but stability. It's average out of the factory and it's going to be average 5 years from now. An enthusiast paste might dry up.

Nope. Actic MX-4 advertises at least 8 years of longevity. And it is the go to paste for many enthusiasts, probably what most use to replace the stock TIM. Noctua NT-H1 has 5 years. But companies don't want to pay 7-8 or 10 times more.

Stock TIM is good enough for most (because they don't know it can be better), and for the rest of us... When we swap TIM to get actual performance from hardware, we loose warranty. Companies are OK for everything that brings them money. And the competition is doing the same. It is win, win, win for them.

 

[Thanos]

10-20C improvement is what you would expect depending on CPU and cooling solution. Whatever the number it is more than enough to prevent thermal throttling.

Power will have a big part to play. The pre-factory overclock of all 8 cores at 4.7GHz is a tall order.

Standard TIM is probably something really cheap. Basically any paste that you replace it with will do better.

If we compare with the worst on the above list then that would mean a 13C improvement would be no better than DC1. This is why a proper reference is required.

They say 13C theoretical value, because results may vary from model to model, and CPU to CPU. So if they say it is 13C better, and someone does not get it there would be complaints. They are covering their behinds.

Theorectical means not proven. Better IMO if something like "Tests have proven up to 13C improvement over TIM 'x' (our previous TIM?) at a CPU power dissipation of 'y' Watts".

Try using google sometimes. https://www.tomshardware.com/reviews/thermal-paste-comparison,5108.html

FYI sometimes isn't enough, I use it many times

I want to see the difference using the Asus hardware and not for instance a comparison done on LGA 775 hardware with an open board (no case) using an IHS. Also Be aware mounting pressure is limited, 15PSI IIRC on a mobile CPU die last time I looked. On top of that that particular link was supposed to have used a thermal diode under the IHS to give Tcase temps except the diode is embedded in the die and the proper way to measure Tcase is with an external thermocouple placed at the geometrical center of the package by milling a channel and recess on top of the IHS for it. Also those DTS sensors at that time have calibration registers to improve accuracy but one needs to know what one is doing or you might end up with something like this