Cooling [Linus] Graphite Thermal Interface Sheet from IC

[rcodi]

30x30mm and 40x40mm in stock now at Amazon for $15 and $20 respectively.

 

[EdZ]

I find myself wondering how a trimmed piece of this stuff would do on a delid. While 35 W/mK is obviously half of what liquid metal provides it will be VASTLY better than Intel's goop.

The goop itself is perfectly fine. The problem is the gap between the top of the die and underside of the IHS:

If you could remove the IHS, leave the Intel goop intact, and re-install the IHS with the gap reduced, you would likely achieve all the gains of delidding.

The cleanliness, containability (ie won't bleed onto your shit and fry your CPU) and need to never replace seems like it could provide a really great middle ground.

Frying your CPU may still be a concern: the pad is conductive, and quite a bit larger than the die. Leaving it to flop off the edge is as bad as having 'liquid metal' TIM on the substrate (so adding insulation around the die would likely be necessary), and cutting it to size would require careful replacement of the IHS to prevent that small rectangle slipping off of the die during re-lidding.
The other concern would be pad thickness: too thin and you have an even worse problem than you started with, as there will no longer by any contact between the die and the IHS. Too thick, and you now have the mounting force of the heatsink going directly to the die (due to it being rigidly coupled to the underside of the IHS) rather than being carried through the IHS mounting to the edge of the substrate.

 

[dumplinknet]

I purchased and tested out this thermal pad.
Conclusion: It made absolutely no difference. Not worth it.

Unit was purchased via Amazon with the Linus TT link.
Purchased the 30x30 which is perfectly sized for the 8700K IHS. Compared to Noctua's NT-H1, on the LP53 heatsink/A9x14 fan combo, the temperature difference was virtually identical.

The installation was straight forward. Clean off the old thermal paste with alcohol, drop in the graphite pad, install the heatsink back in.

I used a pair of tweasers to handle the pad itself. Never touched it with fingers. I noticed that the pad will show an indent imperfection where my tweasers touched the material (like memory foam, but it does not recovery to the original form)- leading me to believe that this is to help "fill-in" those surfaces with micro holes and groves as the heatsink is tighten down. Because of this, reusability is to be questioned.

For science's sake, someone could try unmounting and remounting the pad several times to see if thermals gets worse over time since then the pad would be completely flattened out and would no longer fill in those holes.

Overall, not worth it.

Temps were determined using Prime95 26.6

 

[blubblob]

If the durability holds up that sounds like a great result - for anyone that changes heatsinks often or feels reluctant to handle the pasty kind of thermal interface material.

For the price of 2g of high quality thermal paste it might actually be a decent investment.

 

[rcodi]

If the durability holds up that sounds like a great result - for anyone that changes heatsinks often or feels reluctant to handle the pasty kind of thermal interface material.

For the price of 2g of high quality thermal paste it might actually be a decent investment.

Just got my pad today and swapped for this very reason. I don't plan on changing anything in my primary tower long term aside from GPU so this is perfect although I should be able to reuse it if I did decide to change something. I knew going in that at best it would be on par with the Kryonaut I had on before and at worst a few degrees higher but the difference is irrelevant when I already have plenty of thermal headroom.

 

[tinyitx]

I think the greatest benefit is its taking away the uncertainty of whether or not the thermal paste is spread out evenly, which is good news for beginners.

 

[theGryphon]

I think the greatest benefit is its taking away the uncertainty of whether or not the thermal paste is spread out evenly, which is good news for beginners.

Yes, noone ever claimed these pads will surpass top goops in performance. The value is in ease of use, reusability, consistency and comparability across installations. If I were a tech site, comparing coolers, I wouldn't touch anything else, because these should give more credibility in temp comparisons.

Now, this tech material can be improved in the future, possibly giving a measurable performance lead over goops...

 

[SaperPL]

I think that at this moment this product is a great piece of tech for reviewers and companies that have to do a lot of tests like us.

Sadly I'm still waiting for the availability in Europe - anyone knows where to get those?

 

[tmackerm]

I’d really like to get my hands on one of these. Has anyone from Canada got their hands on them yet, or seen anywhere that ships to Canada?

 

[Tcoo]

I would love to order one for my new build. Love this kind of technology. But they (above Amazon link) don't ship to The Netherlands.
I guess nobody here wants to forward one to me? ;o

 

[Nanook]

I have the 30mm version in 8700K and 6700K builds, and they work fine.

 

[MultiDoc]

Anyone has a link to buy one of these pads in Europe please ?

 

[NateDawg72]

For those who aren't finding a place to get the IC pads where they live it may be worth seeing if you can get Panasonic pads, they are making the pads for IC. Panasonic's "EYG-S series Soft-PGS" is what you would look for. IC Graphite advertises 35 W/m-k while Panasonic's most similar pads state 28 W/m-k. It has worked fine for me though but YMMV.

The one I found that was most similar to IC Graphite is "EYG-S0909ZLX2" which is one of the pads optimized for heat transfer on the Z axis like the IC pads. If you look around and see graphite pads stating huge numbers like 700+ W/m-k stay away from those - that's for the X-Y plane which won't work as well for a thermal paste replacement.

I've been using it on my Ryzen 5 2600 for about a month now. It's a 90mm x 90mm sheet, and I cut roughly a 40mm x 40mm section out for my CPU, and I can use the rest on other builds which is pretty nice.

Here it is on Digikey if that helps with finding it.
https://www.digikey.com/product-det...c-components/EYG-S0909ZLX2/P122034-ND/6575964

https://www.digikey.com/en/product-highlight/p/panasonic/eyg-s-series-soft-pgs

 

[SaperPL]

I got myself some panasonic pads for testing as well. I'm testing it currently on my R7 1700.

I picked "perfectly fitting" EYGS0404ZLMP 36 x 38 mm pads but there was no info anywhere (except for digikey what I found out afterwards) that they actually look like this:

Spoiler: photo

This hole is a disaster for the OCD, but it kind of works anyway, at least it's worth testing.

Here's what you get for $20 before tax - equivalent of IC graphite 40 mm pad:

Spoiler: photo

I got two "pieces" just in case and each of them was a package of 10. Neat.

Here's how it looks on top of Haswell:

Spoiler: photo

I'll try to make some photos on my R7 1700 when I'll get some time to take it apart.

There's few things to note with those Panasonic pads:

1) Take a bigger sheet rather than pads like I did (duh).

2) There are 28 and 20 W/mK variants of those or there are two versions of the datasheet that differ by this number. I couldn't figure out which is which. If the hole in the middle isn't the reason for not performing perfectly as the paste, I'm betting that the ones I got are the 20 W/mK type.

3) Back mounted coolers are a pain to install with this - you either have to glue it to the IHS with a bit of paste (which kind of defeats the purpose) or you have to somehow install this upside down landing the motherboard onto the cooler. I believe there are some variants of those pads that are adhesive from one side, so it would be nice to have something like that simply attached to the cooler.

Performance-wise the ones that I got feel slightly worse (by few degrees) than the paste on the idle/small loads and comparable on full load. I'll have to test it for some time however to get a better feeling on that statement.

Interesting tech. I hope that those IC graphite pads will be available in Europe soon so I can see how much of a difference they make.

I wish they had the variant without the hole in the middle and proper datasheet stating which are 28 W/mK and which are 20W/mK...

 

[NateDawg72]

Glad to see someone else has tried the Panasonic pads

If it helps out others, the full sheet part numbers end in X2, All the others are designed for IGBT modules with holes in them presumably for mounting, see here
https://www.digikey.com/en/product-highlight/p/panasonic/eyg-s-series-soft-pgs

I've been confused about the 28 vs 20 variants, the EYG-S0909ZLX2 I ordered says 28 yet the EYG-S1818ZLX2 which appears to be the exact same thing in a larger sheet says 20

I had the same dilemma with mounting my heatsink. My case had a cutout for the motherboard so I ended up installing the board and sliding the case off the edge of my desk while screwing in the heatsink from below. Definitely makes some coolers harder to install.

I had the same experience performance wise as well, idle and desktop loads had higher temperatures than paste but load temperature was within a few degrees.

 

[NateDawg72]

This last week I was recently working on my PC and switched my video card over to a graphite pad, so my PC is now pasteless. Getting the pad to stay in position on a bare die was a pain and I remembered the conversations here about using a bit of paste to hold it still so I thought I'd share my experience.

If you have a sheet from panasonic, I've found the easiest way to install is to cut a pad a little larger than you need, and tape it to the heatsink with small trips of electrical tape on two edges/corners. No mess and re-usable.

 

[Elerek]

So if I leveled off the pcb of my delidded cpu with the top of the die using some sort of insulating pad (maybe .5 mm thermal pad or something?) and then dropped one of these sheets down, would it outperform a normal delid since with that absurd lateral thermal transfer you're effectively increasing the contact area of the die to the IHS?

 

[Beardedswede]

I purchased and tested out this thermal pad.
Conclusion: It made absolutely no difference. Not worth it.

Unit was purchased via Amazon with the Linus TT link.
Purchased the 30x30 which is perfectly sized for the 8700K IHS. Compared to Noctua's NT-H1, on the LP53 heatsink/A9x14 fan combo, the temperature difference was virtually identical.

The installation was straight forward. Clean off the old thermal paste with alcohol, drop in the graphite pad, install the heatsink back in.

I used a pair of tweasers to handle the pad itself. Never touched it with fingers. I noticed that the pad will show an indent imperfection where my tweasers touched the material (like memory foam, but it does not recovery to the original form)- leading me to believe that this is to help "fill-in" those surfaces with micro holes and groves as the heatsink is tighten down. Because of this, reusability is to be questioned.

For science's sake, someone could try unmounting and remounting the pad several times to see if thermals gets worse over time since then the pad would be completely flattened out and would no longer fill in those holes.

Overall, not worth it.

Temps were determined using Prime95 26.6

What did you exspect? The tests say 2 celcius warmer, but it should spread the heat a bit more, but the hole idea is that you can change cooler more easy and dont need to worry about applying paste wrong.

 

[NateDawg72]

So if I leveled off the pcb of my delidded cpu with the top of the die using some sort of insulating pad (maybe .5 mm thermal pad or something?) and then dropped one of these sheets down, would it outperform a normal delid since with that absurd lateral thermal transfer you're effectively increasing the contact area of the die to the IHS?

It would be worse than a delid with paste. For delidding you should stick to paste or liquid metal. Current graphite pads don't really offer any thermal improvements - just convenience and a reusable thermal interface material.

When trying to apply a sheet across the whole IHS in a delid you have two issues:
First, thermal transfer is mostly going to be limited by the transfer on the z axis directly from the die, so the extra lateral transfer won't help too much.
Second, in order to get good thermal conductivity the graphite pads need a pretty good amount of pressure on them. If you wanted to get decent thermal transfer to the rest of the IHS away from the die, you would need a hard surface around the die that is level with the die. Higher than the die and you lose contact with the die, lower than the die and you don't get any meaningful thermal transfer because of no pressure. An insulating pad would be too soft. Getting a spacer built just right would be a lot of work.