Thermal paste suggestion

[wsarahan]

HI guys

I have a 3770k @ 4.5 and i`m using Mx4 on it

I was thinking to change my thermal paste

I was looking for:

Tuniq Tx2 - Extreme

Tuniq Tx4

Gelid GC Extreme

IC Diamond

Wich one you recommend? Other sugestion?

I will use with a H100 cooler

Thanks

 

[Sohaltang]

I got good results with ic diamond. Plus it's really cheap. It is abrasive if you going to scrub it off many times (it is the hardest mineral).

 

[dma0991]

Local testing by IDC suggests that Noctua NT-H1 is the best. However, changing TIM does not make a huge difference if you already have a decent quality TIM.

 

[Ayah]

IC diamond, to mount and forget due to it's high viscosity. If you demount and remount often, mx4 would be fine.

 

[jaqie]

I personally very much prefer arctic silver's ceramique. It is non conductive electrically, non capacitive, and thermal conductivity is quite good. It also seems to have ZERO creep and bleed.

 

[wsarahan]

The Ic Diamond stains the base too or not?

I bought a week ago the the Liquid Ultra, it will arrive next week, but just later i saw the stains that it makes and the way to apply

I`m a new in this world and i`m a little scared about using the Liquid Ultra, that`s why i`m thinking about the Ic diamond, a couple of guys told me about it

The Diamond is safe right? I mean, it won`t make the satins and the issues that i may have like using the Liquid ultra?

And the way to apply is the same as other pastes if i use the IC diamond? A small ammount in the middle and the H100 on top right?

Sorry about my poor english, i`m from Brazil

Thanks again

 

[Ayah]

IC diamond doesn't stain, just wears the surface down. (think coarse sandpaper)
It's definitely safe, and save for the sandpaper-effect, works like other TIM pastes. And yes, applies well with a large (relative to other pastes) ball-ish thing in the middle using the heatsink pressure to distribute it.

It's not the absolute best performer, but is the best long-term (I'm at 3.5 years, negligible[<2C] performance drop) paste that I know of.

 

[wsarahan]

IC diamond doesn't stain, just wears the surface down. (think coarse sandpaper)
It's definitely safe, and save for the sandpaper-effect, works like other TIM pastes. And yes, applies well with a large (relative to other pastes) ball-ish thing in the middle using the heatsink pressure to distribute it.

It's not the absolute best performer, but is the best long-term (I'm at 3.5 years, negligible[<2C] performance drop) paste that I know of.

Thanks, just one more question about the IC Diamond

Some guys are saying that it will erase the write in the CPU, and mine is still in warranty, is it really will erase the write? Should be a problem?

I`m thinking to buy or the IC Diamond or the Gelid Extreme one

 

[Ayah]

If you mount/demount a ton, it will. If you only do it a couple times, it'll only wear down the writing (it'll fade out)

 

[wsarahan]

I saw good things about this one too, never heard before: Phobya HeGrease Extreme, better than Gelid Extreme?

 

[aarontpx]

Currently using liquid ultra.

 

[Belial88]

Oh god.

Coollaboratory liquid ultra is a huge value, definitely the best at ~$16. It blows other pastes out of the water.

Otherwise, go for a $3 1.5g tube of PK-3 from ebay. A 1.5g tube will last easily more than 20 applications if you use it right, and it's too cheap to ignore. PK-3 is the 2nd best ceramique thermal paste, right behind HeGrease which is really expensive.

Tuniq is crap, Gelid is good, ic diamond is good but too expensive. PK-3, Masscool, or Coollaboratory liquid ultra is really the only way to go.

Thermal paste can result in a MASSIVE reduction in temps, definitely necessary to buy paste. There isn't a huge difference in *modern, ceramique* pastes, but there is a huge difference in a modern ceramique, and a high end TIM like liquid ultra, or a modern ceramique, and an outdated, weak paste like Tuniq, As5, PK-1, etc.

Thermal paste is just like CPUs and GPUs and heatsinks - new ones come out every year and they go out of date very quickly.

Coollaboratory does not stain, people are just morons. You can easily remove it with metal polish. Or, you can use lots and lots of elbow grease (hence why people think it stains, it doesn't, it just sticks well if you dont use metal polish). Be careful about asking for thermal paste advice, you get a lot of people who bought some paste, it's better than the crap they used before, and then they go around saying it's awesome when they have no frame of reference on what's good or not.

I review a lot of thermal pastes, so I've tested tons of pastes out.

 

[colonelciller]

lots of people have opinions on this.
it's probably safe to say that a lot of the presumed positive reviews are the result of a mix of picking up on marketing hype + cognitive bias of one form or another.

this guy did some actual testing... comparing 80!
he also debunks a lot of the misinformation, and puts a lot of the hype into perspective... as well as giving guidance on which paste viscosity to choose for your heatsink based on the characteristics of the CPU interface on your heatsink... and he asseses and debunks common myths regarding TIM application methods

read the article linked (recommended for everyone)

 

[God Mode]

Which paste is the best for not drying out? Initial temps are useless to me. I couldn't care less of a candle that burns brighter than the rest only to die down quicker. I need data on thermal interface longevity.

 

[pcunite]

Which paste is the best for not drying out? Initial temps are useless to me. I couldn't care less of a candle that burns brighter than the rest only to die down quicker. I need data on thermal interface longevity.

I agree. I've been using Arctic Silver 5 and noticed that reapplying after five years helped get the temps back down another 10c again ... to the original starting temp.

Looks like Prolimatech PK-1 is about the best TIM you can easily get.

 

[Ayah]

Which paste is the best for not drying out? Initial temps are useless to me. I couldn't care less of a candle that burns brighter than the rest only to die down quicker. I need data on thermal interface longevity.

IC Diamond is what I've found is the best at longevity and it's why I use it. I've only had it for 4-5 years though. Idle temps increased only about 1-2 degrees for me.

As far as I know, nobody has done comprehensive long term testing comparing constant load versus load/unload cycling (or any kind at all).

So, Coolaboratory has been out for like 6 years, someone should have long term info on it.. But, I suspect most people using it don't just let it be for 6 years.

 

[aigomorla]

http://www.newegg.com/Product/Produc...82E16835242019

PK-1

has been my Aigo Approved Heat paste for about 2 years now.

Aigo Approved meaning:

1. Easy to apply
2. Get near same values on remounts easily (consistency)
3. Very short cure time.. (I HATE ICDIAMOND AND Artic Silver cuz of the cure time, i hate IC diamond even more cuz it scratches anything and everything u apply it on)
3. Get within the top5 ranking in thermal paste.

http://skinneelabs.com/pk-1-asm-cmif/

Here u can see PK-1 under different types of mount... bad / moderate / good

^ graph shows u why its consistent. MX-2 is another favorate of mine with very good consistancy...
I think the best paste is the one that is the easiest to apply.
The easiest to apply paste is the one with the near simular results on multipul mounts.
MX-2 and PK-1 have been VERY VERY good to me during testing when consistent mounts were mandatory.

No tester would use IndigoExtreme to do thermal testing unless it was on indigo itself.
That thing is just wayy too rediculously expensive... if ur gonna bling on TIM, then ur system which goes with it better be worth it.

Ranking right below Indigo, which is cheating... if u look at indigo u'll understand.. its a 1 time use only type of TIM that costs as much as 2 bottles of PK-1.

Coollaboratory does not stain, people are just morons.

no it does something far far worse..
it melts alu.
Its main active is gallium.
Gallium is a big no no with alu.. which is why they warn u on the bottle use only on top of copper.

https://www.youtube.com/watch?v=JHHI2Lk79cY

http://www.overclock.net/t/139742/why-not-to-mix-aluminium-with-liquid-pro

So, Coolaboratory has been out for like 6 years, someone should have long term info on it.. But, I suspect most people using it don't just let it be for 6 years.

no only idiots keep using that.

Once people found out it ate alu it was immediately dropped from our hobby.
I and a lot of testers made it a mission to inform you guys how dangerous that stuff really is.

We saw problems imediate because a lot of heat sinks had ALU in those days....
waterblocks back in those days used ALU...
stock heat sinks today still use ALU...

That stuff was basically pointless to 90% of the people who needed that type of TIM performance.

Its really nasty stuff.... which is why indigo extreme came out which is much safer then that stuff.

However indigo extreme has a "mounting" ritual you must follow which involves cooking your processor @ like 60C+ for a set amount of time so the liquid metal can mold to shape over your IHS.

This is why i said indigo is cheating.... its a 1 time 1 use application that costs as much as 2-3 bottles of the competitors stuff.

 

[Despoiler]

I use MX-4 currently. I have IC Diamond and while good, it is a PITA to work with IMO. It's too thick. You need to warm it up to work with it.

 

[aigomorla]

I use MX-4 currently. I have IC Diamond and while good, it is a PITA to work with IMO. It's too thick. You need to warm it up to work with it.

i never did get into MX-4.
I much prefered the MX-2 over MX-4.

I believe MX-2 was shinetsu g351... while MX-4 was something else.

Shinetsu is another GREAT TIM.


But guys if ur looking for TIM... PK-1

Trust me on it... you wont look back.

 

[pcunite]

But guys, if you're looking for TIM ... PK-1

Trust me on it ... you wont look back.

Thanks for the info. It does test very well.

 

[blastingcap]

Thanks for the info. It does test very well.

I can vouch for PK-1 and MX-2 having personally used them. Both are better than AS5 which takes forever to cure and is electrically capactive and costs more. MX-4 is good too.

 

[sLAMAZAUR]

Noctua NT-H1 ftw. from the ones you listed I'd go with IC Diamond

 

[Ajay]

http://www.newegg.com/Product/Produc...82E16835242019

PK-1

has been my Aigo Approved Heat paste for about 2 years now.

Aigo Approved meaning:

1. Easy to apply
2. Get near same values on remounts easily (consistency)
3. Very short cure time.. (I HATE ICDIAMOND AND Artic Silver cuz of the cure time, i hate IC diamond even more cuz it scratches anything and everything u apply it on)
3. Get within the top5 ranking in thermal paste.

http://skinneelabs.com/pk-1-asm-cmif/

Here u can see PK-1 under different types of mount... bad / moderate / good

^ graph shows u why its consistent. MX-2 is another favorate of mine with very good consistancy...
I think the best paste is the one that is the easiest to apply.
The easiest to apply paste is the one with the near simular results on multipul mounts.
MX-2 and PK-1 have been VERY VERY good to me during testing when consistent mounts were mandatory.

No tester would use IndigoExtreme to do thermal testing unless it was on indigo itself.
That thing is just wayy too rediculously expensive... if ur gonna bling on TIM, then ur system which goes with it better be worth it.

Ranking right below Indigo, which is cheating... if u look at indigo u'll understand.. its a 1 time use only type of TIM that costs as much as 2 bottles of PK-1.



no it does something far far worse..
it melts alu.
Its main active is gallium.
Gallium is a big no no with alu.. which is why they warn u on the bottle use only on top of copper.

https://www.youtube.com/watch?v=JHHI2Lk79cY

http://www.overclock.net/t/139742/why-not-to-mix-aluminium-with-liquid-pro



no only idiots keep using that.

Once people found out it ate alu it was immediately dropped from our hobby.
I and a lot of testers made it a mission to inform you guys how dangerous that stuff really is.

We saw problems imediate because a lot of heat sinks had ALU in those days....
waterblocks back in those days used ALU...
stock heat sinks today still use ALU...

That stuff was basically pointless to 90% of the people who needed that type of TIM performance.

Its really nasty stuff.... which is why indigo extreme came out which is much safer then that stuff.

Thanks for the great info! Even though I stick to copper, there might be Al on some part of the HS, and that would be bad news long term. Is there any problem with pump out w/PK-1?

 

[aigomorla]

Thanks for the great info! Even though I stick to copper, there might be Al on some part of the HS, and that would be bad news long term. Is there any problem with pump out w/PK-1?

pump out?

Do you mean as u apply the heat sink on top and the residue which fills out?

PK-1 uses the universal method...

a Rice grain line in the middle of the IHS, and then apply pressure.


If your talking about getting the paste out.
No its not as bad as ICDiamond which u need to warm the bottle sometimes to get the stuff out.

 

[Ajay]

pump out?

Do you mean as u apply the heat sink on top and the residue which fills out?

PK-1 uses the universal method...

a Rice grain line in the middle of the IHS, and then apply pressure.


If your talking about getting the paste out.
No its not as bad as ICDiamond which u need to warm the bottle sometimes to get the stuff out.

No. Differential thermal expansion of the CPU package and the HS lead to the TIM being slowly pumped out of the contact patch, or at least, the more aqueous portion being pumped out leaving the solids behind. This can be most easily seen in GPUs which use 'white paste' and when you clean and remount it after a year or two, all you see is a dry hard white compound between the GPU and the HS.