Graphite thermal pad, a solution for uneven core temperatures?

This morning i came across this product.
https://www.innovationcooling.com/initial-desktop-tests-ic-graphite-vs-gelid-gc-extreme/

A graphite based thermal pad which is ment as an alternative to thermal paste. Thanks to this design pump out is less of an issue and because of the thickness in theory the uneven core temperatures might be solved with this some laptop users.

It also does not dry out!

I might want to try this, not sure when i can get my hands on it though.

 

They also have a laptop test with this stuff. Looks promising.

https://www.innovationcooling.com/ic-graphite-notebook-testing/

 

http://forum.notebookreview.com/thr...mal-pad-available-for-test-and-review.815439/

 

Papusan just gave me this thread link as well. going to read into it.

 

This looks very interesting. Only ever needing one repaste (repad in this case) would be great. Is this the future?

Sent from my Pixel XL using Tapatalk

 

Teh results in the main thread seem to vary. But what I would like to find out is if it lowers the peak temepratures of users here with higher temperatures differences between cores. For example if 2 cores are 70c and the 2 others 80c with normal paste, but with these pads for ecample 75c on all cores it still would be beneficial compared to normal paste.

 

I am one of the testers, it is an inferior product compared to liquid metal compounds.it delivers about formula 7/ICD7 performance but doesn’t dry out supposedly

 

Hmm but in that case its a pretty good solution I suppose for people that have uneven temps but do not want to risk it with liquid metal?

 

I'd imagine most people don't use Liquid metal, I for one am hesitant to risk it with a laptop I carry around in my backpack, so for people like me it could be useful. I currently use kryonaut

Sent from my Pixel XL using Tapatalk

 

Simple answer nope. And it does not perform as well as it does in desktops, because of the mounting pressure.

 

I've carried my M18x to and from school for 3 years with CLU and never had a problem. CLU likes to stick to itself and bonds to metal.....not exactly going to fly out of place.

For the brave, they may try these:
https://www.arrow.com/en/products/eyg-s0912zlgd/panasonic

 

Do you have any numbers? because they promote a laptop case on their website but that one houses a fairly old CPU.

 

See my results:
http://forum.notebookreview.com/thr...r-test-and-review.815439/page-8#post-10713397

 

Hmm interesting. But CLU is ofcourse the ultimate to compare it with. Hopefully someone does a few comparisons vs kryonaut, icd7 and similar pastes.

I use Kryonaut now, but it is slowly degrading. So my temps went 3c up after 3 months of use.

 

It's not worth using conventional paste. You should seriously consider using CLU or the likes and forget about repasting ever again.

 

I worry tha tthe slight uneven-ness of my heatsink will cause issues in the long run. I have a 3~4c core differential with a fresh paste.

 

I thought you had resolved that with changing the size of thermal pads

 

@Raidriar Did quite a few tests. And ill be doing my own in the AW17 R4 later this week. But im not expecting anything different.

EDIT: Whoops he already replied

 

How thick was the pad you recieved? I found a different one with similar properties, just trying to compare. The one in question I am looking at is 2mm, was the sample you received the same thickness or less?

 

0.25mm

 

I never changed my pads but it could on other units. But out of the box I had 2/3C core temp differences, I considered this within margins. But Kryonaut made this larger. Probably because of the imperfect heatsink. Still need to repad.

Edit: i did on others people machines though and in that case it helped a lot. But they had older revision heatsinks.

 

That's disappointing as it suggests this may be the same PGS-S I was looking at as an alternative.

 

I'd say 0.5mm or even less

 

Curious about your findings. I will keep an eye on it.

 

Well a single sheet of paper folded in half is .1mm so was it really that thick?

 

it was extremely thin. I don't have a caliper on hand to measure for you

 

No worries, the reason I ask is because of my post here:
http://forum.notebookreview.com/thr...-test-and-review.815439/page-16#post-10720521

What I found looks similar to what you were given but at .2mm that makes it pretty thick and my hope was that it would be enough to apply adequate pressure and was trying to compare notes.

 

Read this PDF (it has all sorts of good information from what I would consider an scientific and reputable source):

https://www.google.com/url?sa=t&sou...FjADegQIABAB&usg=AOvVaw3WMWJCvGYDYnmT5h7lIu_N

 

Thanks for sharing!

Indium conductivity (one of component of CLU) seems to be more good than a soldered interface (0.83W/cm VS 0.51).

I'm one of lucky tester of IC Graphite thermal pad proposed here http://forum.notebookreview.com/thr...mal-pad-available-for-test-and-review.815439/

I received them last week, So many Thanks to @Innovation cooling for materials and shipment (a fast ship by UPS tracked, from US to Italy)

I received various pads of 2 different size, I 'm going to test them soon, both over my CPU/GPU on MSI 16l13 aka Tornado F5 by EUROCOM, that got a 7700K on a Bitspower's cpu IHS (Skylake ver.) that maybe help a bit with the common pressure issue, and a GTX 1080 with 150W vBios
(I will test them against the king of TIM aka CLU and VS some other like Phobya Nanogrease Extreme and ArcticCooling MX4 that I got already at home)

 

I have gotten a couple of the pads. 30mm and 40mm versions. Between the lapping, the shim soldering and the Rockitcool IHS on my 8700k . The temps were semi decent. I can say this about them though. The higher the mounting pressure, the better the temps. I ordered a butt-ton of shouldered M3 bolts with different springs for pressures, e-clips, various lengths of M3 bolts both hex and socket heads during my shimming process. So I have been able to test different pressures as opposed to the stock heatsink shoulderd screws that came with the p870tm1-g . Once a new heatsink arrives (hopefully soon) , I plan on testing the pad a bit more without a shim in place to see if the stock sink can benefit from the pad with different pressure applied. Also got the DIE guard finally printed and tested so awaiting for some eSUn ABS+ to be delivered so I can make the next one with it (temprature related shennanigans) then i can fully test the heatsink I have soldered the IHS to and for a bare DIE approach with proper pressures of holding the cpu in place.

 

Here's some results with my 15R3
Both GPU and CPU are undervolted (CPU -150mV), ambient temp is about 28°C

First with OCCT

Before: Kryonaute, one month old repast.


After: Graphite pads


They didn't do miracles. But being able to tweak the VRM's pads and tighten/loose the heatsink's screws without caring about cleaning and repasting each time was quite convenient.
Had To try 4 times before i could find the perfect balance.

But performance wise, if you compare GPU's temp with 5mn of Ungine Heaven you'll be disappointed :

Then Unigine Heaven

Before: Kryonaute, one month old repast.


After: Graphite pads


Sure CPU's temps are nicer, But GPU's temp isn't. 2°C above.

I wrote a Graphite pads review on P4G's French forum if interested.

 

I think using the graphic pad on CPU only is the best way to do it. The CPU receives the benefit of extra pressure. The GPU side doesnt need that.

 

It wasn't really the extra pressure who helped. Graphite pads are pressure sensitive too.

First try, cores 0&2 were 90°C+ and cores 1&3 where 67°C ! Not enough pressure at the top.
2nd try, I swapped the 0.5mm CPU VRM's pads with 0.15mm and tightened only the top screws. Results: cores 0&2 were 70-72°C and cores 1&3 were 76-82°C. Not enough pressure at the bottom.
3rd try, I tightened a little more the bottom screws. Results: Cores 0 and 1 were hotter than the 2 others. Not enough pressure to the left side.
4th and last try, I tightened fully the GPU's bottom screws and voilà. The pressure was even.

 

Sounds like the memory thermal pad next to the CPU heatsink is too thick. This was the same pattern that I experienced. My temeprature imbalance wasnt the usual Core 0 and core 2 hotter than Core 1 and 3. But Core 0,1 where the highest and core 2 and 3 where the lowest. I changed the memory chip pads and that fixed my issues.

 

Biggest problem is the CPU BGA slug is convex :/
and it doesn't help when the heatsink is also convex too.

Also the VRM pads and VRMs at the top make this hard to balance