Absolute best Thermal Compound

Coolermaster says on their own website this is "non-curing" thermal paste. The only manufacturer that says "you need a lot of time for the paste to get best performance" is Arctic Silver.

I also found this:
http://forum.notebookreview.com/threads/cm-mastergel-maker-nano-keeps-drying-up.827434/
That person had an issue with uneven pressure but it is obvious this paste is unsuitable for notebooks since it cannot compensate for this. I have also seen big core difference when I also used it, when I applied Gelid the core difference was much less.

 

Noctua and GC-Extreme, end of story. Anything else non-conductive dries up with in a month

 

I'm sure bro @Mr. Fox ( Mr. Fox @ HWBOT) do not agree with this http://forum.notebookreview.com/thr...ers-welcome-too.810490/page-229#post-10727319

http://forum.notebookreview.com/thr...-owners-lounge.826831/page-1196#post-10917287

http://forum.notebookreview.com/thr...ers-welcome-too.810490/page-230#post-10727500

And I know the newer Noctua NT-H2 ain't so much bette.

1) Conductonaut
---------------------------------------------------
2) Phobya Nano grease Extreme
3) ICD (best with awful fit from HS on die/IHS).

Regarding Liquid metal we have also Phobya Liquid metal - Coollaboratory LIQUID EXTREME. Not sure about the new Liquid metal from Coollaboratory. They screwed up the recipe on older Liquid Ultra, so none know how the coming batches will be. Need to be tested.

We have also the not so known Liquid metal from Alphacool. Alphacool ice frost Xtreme liquid metal. Probably on license from Aquatuning GmbH ~ Phobya.com (Exactly same Thermal conductivity numbers: 40W/mK as Phobya Liquid Metal). No one has mentioned this LM before as far as I know. So here it is

 

My ICD repastes of up to 10 yrs haven't dried out and still perform as good as day 1.

 

ICD definitely does not dry up but after scratching the surface of CPU with it (before any idiot will say "you did not apply it right or did not remove it right" - I have been respasting for over 10 years, this is the only thermal paste that did physical damage) and after having VERY hard time spreading it out (I want to always be sure the paste has completely covered the surface, the only way to do this is to spread it manually) - I am never touching this again.
Especially after witnessing:
http://remixedcat.blogspot.com/2013/07/innovation-cooling-ceo-throws-tantrum.html

 

Even Grizzly messed up one or two batches for their Kryonaut paste. And you won't see degraded cooling performance with some cosmetic scratces. What is better... A shiny die running hotter and hotter as time go and you'll need repaste or put it on and forget it?

 

Take a good look on Google, see for how long the reports of ICD causing scratches were appearing. Last time I personally used ICD was this year (I can show you a proof of purchase), it left significant scratches which can be felt with the nail on the surface of CPU not covered by heatspreader. The issue is more than "one or two batches" ;-)

The scratches which you can feel are not just cosmetic, they are permanent until you polish them off with some compound (which no user should be doing). And they definitely can affect thermal performance. Here is a fact for you: the best thermal conductivity is when you have direct metal-to-metal contact. A scratch is basically a gap which will be filled with paste, the paste will have worse thermal conductivity than direct metal-to-metal contact. A few scratches might not produce enough temperature difference, but the more you will add them - the more you will see the difference. This a pretty logical conclusion based on scientific facts, not random assumptions ;-) You can even do your own experiment - polish the surface of, say, heatsink to a mirror-like state, then test how well this will work (for most objective results you do not use any thermal paste so you won't have to factor in the quality of thermal paste and quality of application), then scratch it with multiple scratches and do a re-test. You will absolutely get a measurable difference after you'll put enough scratches.

There is also a possibility that when you will return your laptop for repair - the seller might notice scratches on the surface of CPU and void the warranty by claiming "user damage". It's highly unlikely but I would rather completely avoid this possibility if I can ;-)

 

One of the biggest users of ICD and as well test several brand paste the last 8-10 years is bro @Mr. Fox. He have done several re-paste with ICD the later years for his overclocking passion. Not that this was needed but he always hunt after best result every time. An you won't see degradation in cooling performance if properly done. If you run into this... Maybe change how you clean it of. Search up and see if you find pict/video of die after use of ICD. I'm sure you'll find it. Or maybe bro Fox will show you some pict

For my own part I have always run after best possible performance and best possible temps. You'll find them if you search. I know what works on my own hardware.


And I don't re-paste several time a year as some others <must do due pumping out>. Not the best thermal paste is doomed to do so

 

@Papusan Noctua NT-H2 vs Kryonaut. Tested both on my P775TM1-G. Noctua is around 0.5- 1°C worse than TG.

Both tests done on 22°C ambient temp room. I always apply too much paste so I don't know if I done it properly what kind of results I may get.

Sent from my SM-G970F using Tapatalk

 

Look at my post about lapping. Be sure you can see that the paste being spread out evenly. Soft toothpaste works well if you don't have pressure pape(they are damn expencive). Or don't want waste expencive thermal paste. And don't forget the Clevo Cpu bracket lottery
http://forum.notebookreview.com/threads/clevo-overclockers-lounge.788975/page-1306#post-10560808

 

It scratched the die because you spread it out, you do not need to do this if you have enough contact pressure as that will spread it out for you. I've been using it for years and i've never scratched a die with it.

 

1) DO NOT TRY TO SPREAD IT. It's not designed for that. You can do a thick line, or a ball, depending on die shape, and the heat from the CPU will VERY QUICKLY soften the paste and the heatsink pressure alone will spread it, and the viscosity of the paste will drive out air and fill any gaps and scratches. Since the paste is non-conductive electrically, and suffers little from thickness, you can add extra paste, and just let it run off the sides.

2) Excessive rubbing can scratch a die surface. Ergo: DO NOT DO EXCESSIVE SCRUBBING. If the paste won't lift off with isopropyl alcohol and a lint-free surface, like a coffee filter, then use a solvent with acetone. I use Goof-Off, and a single drop works miracles. Then wash with alcohol to remove residue.

I've cleaned many units and never scratched a surface with ICD7 because I understand chemistry, and I never spread the paste.

3) I've redone units at Asus (desktop and laptop) with ICD7, and never had a complaint about the paste failing. In fact, Gary Key (Manager of Technical Marketing for North America) came by one day and told me that because of my notebookreview.com thread on my repaste job on a G73JH, and subsequent 23C temp drop from it, and how it lasted, that Asus changed from using their stock grey paste and switched to another product. (Not ICD, but still better)

4) The problem with quoting forum postings as proof of your argument is that your sources may not be correct, and should always be taken skeptically. My experience with the paste comes from my professional experience working on 100's of systems, plus my own personal DIY stuff. I miss doing LN2 overclocking.

I love the paste for notebook systems, due to heatsink lottery. (My G73JH heatsink had a gash in the copper face due to a screwdriver wedged into a previous install. Yes, heatsinks get reused.)

 

I am happy for him. Not sure why you mention him, though. Just because he likes ICD will not change the FACT that it scratched the CPU on my laptop. Or scratched CPUs on other people's systems (Google can find them). And I don't want this happening on another laptops, even if they are not my own.

I know how to clean. I use Acetone-based cleaner, it cleans off any thermal paste better than alcohol and I always gently wipe out the remains with soft tissue, regardless which paste I used. And cleaning is irrelevant if the scratch will appear during application.
Also, you fail to read what I typed above. Once again, I have NEVER scratched anything by using dozens of other thermal pastes. And I do not repaste only my own PCs.

So do I. For me, Gelid worked best out of other thermal pastes. Whether it works for other people as well and whether they will also use it - I don't really care, I do not receive any money from Gelid, all I can do is to recommend it based on my personal experience and it is up to others if they want to try it out ;-) Same goes for warning other users about physical damage caused by IC Diamond.

ONCE AGAIN, just as I said before, I do NEED to spread it out manually, this is the only way for me to know for sure whether I applied enough and to prevent excessive over-application (yes, I am aware that excessive paste will not damage anything if it is not conductive). If you prefer to gamble and just put a blob in the middle or draw crosses and HOPE you applied enough - that's your preference, doesn't mean this is the best way for everyone.
This is also irrelevant if you get something like the other user at Techpowerup forums:
https://www.hardwarebbq.com/ic-diam...e-hand-thermal-pastes-damaging-cpu-gpu-cores/
Look at the photos, they show not only scratches but pitting which is a result of heatsink pressing on some hard particles within the paste. This can happen regardless of how you spread or clean the paste.

I am glad that you were. However, it does not change the FACT that other people were not so lucky with it and it either scratched or pitted the surface of the CPU and GPU (I wonder how many times will I have to repeat this until people will be able to comprehend this?).

I am sorry but I prefer to trust multiple sources reporting same issue instead of trusting, say, an experience of a representative of the company whose job is to promote the product ;-) Especially if the experience of other people in those sources is very similar to my own experience.

 

Aka confirmation bias.

 

I don't work for ICD. I said I worked for Asus. My banner was a TY from IC owner for demonstrating th ebenefits of his product in an open and factual manner.

 

I'm sure you wanted to hear how it works from an overclockers world. Not from Average Joe

Once again... Apply new ICD after properly cleaning and you'll see same temp as last time.

Thats ok. What works for me doesn't mean it is best for you. Test it on own hardware is what <all should do>

 

Sure, you can call it this way. That doesn't change the fact that whatever I mentioned can happen

 

Arctic silver never really worked for me! My ambient temps are usually higher so mine needed a curing time of 1 week and its been 4 years w/o repaste.

 

Why don't you guys just make your own liquid metal galinstan?
68% Ga, 22% In, and 10% Sn

You could increase Indium to 30% too for more w/mk.

https://www.amazon.com/gp/product/B01F5SB94C/
https://www.amazon.com/gp/product/B01BN7QEOE/
https://www.amazon.com/gp/product/B005DRQRZM/

https://www.amazon.com/Premium-Graphite-Crucibles-Refining-Aluminum/dp/B01N9TNLSG/ for the smelting.

Then grab some small needle syringes.

https://www.reddit.com/r/watercooli...yd_liquid_metal_thermal_paste_and_so_can_you/

 

I would use liquid metal, I just don't want to lose the warranty - I don't know how Dell will react to seeing the Gallium diffused into heatsink.

 

Gallium only diffuses into the very top layer of the copper, and this gallium can be wiped out without any damage to the heatsink whatsoever with 5000 grit sandpaper.

 

Yea, I know, I just don't want to bother with it. Gelid worked good enough for me. No thermal throttling in games, only time I still see high temps is when using Forza Horizon 4 benchmark.

 

It's only about theory of relativity. (Hope my middle school physics teacher won't slap me). I have tested out MX-4, kyronaut, GCE and NT-H1 on my mining rig (2x GTX 1080 Ti FE and 4x1070). That's like 84C for 1080 Ti FE for a continuously 3 months or so. Very straight forward comparison.

And previously I have had HD 7970/ R9 nano with MX-4 for longer and a few years prior when I was playing a lot around with my thinkpad T61. MX-4 was good for P8700 which can run under 65C. But when it comes to T9900/X9100/QX9300 it gets cooked up very easily.

I purchased my thermal pastes directly from newegg to avoid possible chinese fakes.

I have no doubt that things like liquid metal and ICD 7 would perform much better than the bestest thermal paste. But they are a bit controversial to begin with.

Also just FYI, the Chinese are making some insanely cheap Liquid metal. Gallium and Indium alloy is not rocket science

 

Really? You have it all figured out don't you?

I've tested them all, literally.

Best Liquid Metal = Thermal Grizzly Conductonaut
Best non metal = Phobya NanoGrease Extreme.

been running them on every laptop and their performance never deteriorates and they *don't* dry out

 

Never tried Phobya NanoGrease Extreme. Sadly I cannot buy it directly through Amazon, only through third-party resellers there. It also has a pretty low viscosity similar to MX-4, so I wonder how long it will actually last in laptops:
https://www.tomshardware.com/reviews/thermal-paste-comparison,5108-12.html
Maybe I should try Kingpin kPx instead, though they don't list the thermal conductivity for it.

 

Not rocket science but you should still know a bit about metal chemistry (from 14.25). As well some info about Pump out effect vs Notebook cooling

You should read next page... Results: Usability

"The lower the paste’s viscosity, the easier it is easy to handle. However, there are even more factors to consider. For instance, some pastes are stringy like cheese on hot spaghetti. This is a subjective rating, of course, but it often says more than a simple viscosity rating"

Kingpin kPx was tested by one of us in Clevo Overclock thread... Not the best for notebooks.

 

I did a search for Kingpin kPx, couldn't find any results of anyone testing it on these forums.

 

Kingpin can be written several ways and add in more text for the search help as well... K|ngp|n Cooling Thermal Paste

With much higher clocks then I expect that the differences will increase. A notebook with less steady cooling design than the big boy Clevo P870 won't benefit much from this paste vs. the others.

 

Thanks for finding it. These are actually pretty good for conventional thermal paste. I definitely need to try a paste with higher viscosity, today I noticed much higher temperatures in Prime95 compared to my last repaste with Gelid. I disassembled laptop again, removed heatsink and there was barely any thermal paste left between heatsink and CPU. Looks like Gelid is also not viscous enough, just like previous pastes I have tried. It gave me great results at the beginning but now it's back to this in Prime95, with large temp difference between cores, this image I took today before repaste:


Here is the image after today's repaste. Core temp difference is less than 10c again:


Long story short - Gelid GC Extreme also sucks since it does not last long. Basically I have 3 choices now: go back to IC Diamond and risk scratching CPU and GPU, buy Kingpin kPx or go with liquid metal (it does not seem to deteriorate that fast).

 

http://forum.notebookreview.com/thr...6-owners-lounge.797128/page-842#post-10614277

 

Yea, I've read that, but according to the reviews which show viscosity, Phobya Nanogrease is also not thick enough, and I don't want to keep wasting money on testing these, I already have too much of various conventional pastes ;-) I'll give Kingpin kPx a try, if it will be stable for long time - I'll just keep using it. If not then I'll just go with liquid metal and maybe use kPx as a seal for it.

It would be great if this forum section had a sticky post where people could share the longevity of the paste they have tried, with temperature screenshots to show the performance over time.

 

Welp that's among the paste I have tried. Haven't tried that phobya one. The most recent non-conductive I have tried is the kyronaut, which turned out to be overpriced and hyped.
What is the ingredients of Phobya paste? it sounds like tiny particles of diamonds?

 

Grab a 800 grit sand paper 3M or any local brands and just lightly rub the heatsink in one direction. Rub or lap the heatsink where the paste hasn't been dried up and don't lap too much since laptop heatsink lack pressure and too much lapping will make thermal even worse than before/stock!.

 

Product Info
NanoGrease Extreme 16W/mk thermal grease - data sheet

 

Is it enough? I’m sure I could swim in it, HeHe

Spoiler

Liquid Metal to remove Indium solder

 

Prime95 small FFT temps:


Hopefully Conductonaut will last longer than conventional pastes. Also, fun fact about Dell G7: there are no exposed metal connectors or resistors on surface of CPU and GPU, so even if you will apply too much liquid metal - there should not be any damage as long as it won't leak past the area of CPU and GPU.

 

That would be end of my PC forever on a bumpy road!

 

Interesting you say the NanoGrease extreme works the best in laptops. What I hear the most is that people want a "thicker" paste for laptops due to uneven contact and low mounting pressure. Does Nanogrease extreme work better then a "Thick" paste in that regard? Im curious as to how much better does nano grease does and what would be second best. Most people dont do laptop thermal paste testing, and just about every thermal paste test there is often points it to kryonaut. For laptops I believe its Gelid extreme in regards to a thick paste, but I have never seen any comparison of that with kyronaut, and phobya nanogrease extreme. Im curious how them compare.

 

Where do you buy Kingpin paste?

 

There is only one place to buy.
https://kingpincooling.com/products/kpx-high-preformance-thermal-compound-3g

 

the shipping costs more than the paste itself :/

 

It probably wont work good for laptops, and doesnt even look thicker then kryonaut after seeing a video of it, and if anything its more watery. From what I have heard, its sole point is for Liquid nitrogen cooling as most paste will freeze up while this one wont. Looks like DX1 is the way to go for paste.

 

You're wrong. Kingpin kPx has much higher viscosity than pastes like Kryonaut. Meaning it is much more thick.
https://www.tomshardware.com/reviews/thermal-paste-comparison,5108-12.html

@Temp1234453 If it costs too much - try other paste like IC Diamond. It also has high viscosity and will last very long.

 

I do have to laugh at that "test" for viscosity. It's a spread test and nothing more, it doesn't tell you what the viscosity really is or what it is like at operating temperature. I really thought I'd see some sort of science or something behind it. They also rated ICD 24 as having one more arbitrary viscosity unit than ICD 7, lol.

 

I agree that their numbers are arbitrary but it still is pretty accurate since you can find viscosity numbers for some pastes and compare them to their arbitrary numbers in terms of which paste is more viscous than the other. Sadly Kingpin and some others do not publish the actual numbers. As for difference between ICD pastes - it is possible that they got different batches, one was older and more difficult to spread and other was newer and easier to spread.