What the best thermal paste in 2019?

[Raven Rampkin]

Love the sheer number of chairs flying in the audience What a surprise (NOT).
Therefore,

"What the best thermal paste in 2019?"
Whatever makes your heart pound with joy
Gotta catch try 'em all!

 

[Bill_Bright]

I wanted to step out of this thread but since I was quoted directly...

Sry. 200 hours is not an extreme example but the actual recommendation from Artic Silver application instruction pdf. You can find the link in one of my previous posts.

Arctic Silver 5

Then the compound thickens slightly over the next 50 to 200 hours of use to its final consistency designed for long-term stability.

"Over the next 50 to 200 hours". The application instruction you linked to that you even quoted says the drop will occur "over the break-in period". It does not say it will take the full 200 hours every time. 200 is the maximum, not standard amount of time.

If someone tells you something will happen "over night", does that mean it won't happen until 1 minute before the sun rises? No. If they say "over the next 1 to 3 days", does that mean not until 71 hours and 59 minutes have passed?

The issue is that some people are more enthusiast and seek that 2c degree drop in temps and it may matter to them.

No, that's not the issue. That's not an issue at all. That's just an opinion, a "desire" based on preference, not fact. As noted in my signature, you are entitled to your opinion, but not your own facts.

You quoted me. Did you read what you quoted?

I don't have a problem with someone wanting those 2°C. But wanting and "needing" are two different things you and others don't seem to understand.

If you are competing in a contest to see who can achieve the lowest possible temps, then okay, you may "need" those extra 2°C. Otherwise, it is just about bragging rights, or to compensate for a failure to properly setup case cooling or a failure to apply the TIM properly in the first place.

 

[Sithaer]

Me and my bro use MX 4 for years now since its cheap and good enough for our standard use,don't see the point in changing it.

 

[Vayra86]

I'm not sure what the best thermal paste is nowadays, I'm still using MX2 in 2019. I love the stuff and don't see any reason to change. Considering I have about 14g sitting on my bench...I'm stuck with it for quite some time. .

I enjoyed Ford's post in #152. I personally use the grain of rice method. For me, it gives the most uniform pattern without air pockets. Although...on GPU's I spread a thin layer across the whole of the processor. The reason is...I don't think the clamping pressure of the coolers on these is as high as a regular normal desktop processor cooler and I've run into overheating problems when using the rice method on GPU's. Now...admittedly, I haven't tested this theory in probably over a decade...so "things" could have changed since then. Either way...its been working for me and I'll probably continue to do it that way until it doesn't.

Another small observation though...

Having taken apart many many many OEM computers over the last 4 decades...the most air bubbles &/or pockets that I have witnessed came from factory oem computers where the thermal paste was applied over the whole processor before assembly. Just as in Ford's photos. Hence my comment.

What I find strange is that this behavior doesn't repeat itself when I pull off my GPU coolers years later. Despite using what I consider to be the "inferior method"...I never see air pockets? All of these years, I chalked it up to clamping pressure as mentioned.

Although...the truth is...I don't know.

Best,

Liquid Cool

Isn't GPU cooling done with different pastes altogether? That's the only explanation, really. More often than not the gunk is not even just on the die, but also over the ICs around it.

I don't have a problem with someone wanting those 2°C. But wanting and "needing" are two different things you and others don't seem to understand.

Bill, you are a wise man.

 

[Bones]

All I can add is the curing time stated as being 200 hrs for AS5 is a ballpark figure that's reasonable, it's true the time will vary between builds, no question of that but 200 is a good number related to when you can expect it to be fully cured, as in if it isn't by then it never will be.

I don't have a problem with someone wanting those 2°C. But wanting and "needing" are two different things you and others don't seem to understand.

If you are competing in a contest to see who can achieve the lowest possible temps, then okay, you may "need" those extra 2°C. Otherwise, it is just about bragging rights, or to compensate for a failure to properly setup case cooling or a failure to apply the TIM properly in the first place.

Yes, 2c can make all the difference in the world running a comp - I know that firsthand.

However alot of folks using different, more exotic TIM's isn't anything to do with bragging rights, it's they see others using it and "Because those guys are using it I should too".

"Monkey see, monkey do" mentality.

That being the mindset of some which isn't the smartest way to go and in some cases definitely NOT the way to go.

An example of that is the use of liquid metal TIM.
It does work, no question but it's also known over time it works on surrounding materials causing damage. While the shorter term benefits are good the longer term effects are destructive to just about everything it's in contact with.
However many don't worry about it since they upgrade components often, passing this issue along to the next guy to use the hardware - Just hope you're not one of those guys.
TBF if it's working for you be it liquid metal, AS5, Kryonaut or whatever else I don't see why you should change and bottom line is it's your money and hardware anyway, do as you please.
You've every right to.

As compensating for a poorly done install I doubt 2c would make enough difference to offset the screwup - In a few cases it just might do the trick but I woudn't hold my breath over it.

 

[freeagent]

I can’t believe how much people dislike as5 lol. It’s like they shredded a box of kittens in front of you or something. I mean really, at best there is only a few c between them all, and most of it comes down to your application technique.

 

[Deleted member 185158]

I wanted to step out of this thread but since I was quoted directly...

Arctic Silver 5 "Over the next 50 to 200 hours". The application instruction you linked to that you even quoted says the drop will occur "over the break-in period". It does not say it will take the full 200 hours every time. 200 is the maximum, not standard amount of time.

If someone tells you something will happen "over night", does that mean it won't happen until 1 minute before the sun rises? No. If they say "over the next 1 to 3 days", does that mean not until 71 hours and 59 minutes have passed?
No, that's not the issue. That's not an issue at all. That's just an opinion, a "desire" based on preference, not fact. As noted in my signature, you are entitled to your opinion, but not your own facts.

You quoted me. Did you read what you quoted?

I don't have a problem with someone wanting those 2°C. But wanting and "needing" are two different things you and others don't seem to understand.

If you are competing in a contest to see who can achieve the lowest possible temps, then okay, you may "need" those extra 2°C. Otherwise, it is just about bragging rights, or to compensate for a failure to properly setup case cooling or a failure to apply the TIM properly in the first place.

Wanting and needing are one in the same when comes to competition. Not following this as 2 seperate things. I want and need cooler temps simultaneously.

A guy spending 500$ on a custom loop and using a sub par paste is just defeating the idea of cooler temps.

A guy going from air cooling to AIO also using sub par paste is defeating the purpose of the seek for better temps.

So the quality of paste should be just as important as a want or need as buying an after market air cooler which may only deliver 2c drop in temps with a good paste an additional 2c drop in temps, 4c drop total may net that extra 100mhz oc desired by the overclocked.

So in reality, everyone here is an enthusiast to some degree, or you may find them at the Dell forums instead.
Otherwise, none of us would be having this conversation.

 

[Vayra86]

Wanting and needing are one in the same when comes to competition. Not following this as 2 seperate things. I want and need cooler temps simultaneously.

A guy spending 500$ on a custom loop and using a sub par paste is just defeating the idea of cooler temps.

A guy going from air cooling to AIO also using sub par paste is defeating the purpose of the seek for better temps.

So the quality of paste should be just as important as a want or need as buying an after market air cooler which may only deliver 2c drop in temps with a good paste an additional 2c drop in temps, 4c drop total may net that extra 100mhz oc desired by the overclocked.

So in reality, everyone here is an enthusiast to some degree, or you may find them at the Dell forums instead.
Otherwise, none of us would be having this conversation.

Need to be an enthusiast to 2 degrees or you won't need the best paste

Jokes aside, you're very right, but don't mistake enthusiast with knowledgeable... There is a staggering amount of people that buy thousands of bucks worth of half useful parts and assemble it like they play with Duplo. And when it comes to paste, that is a critical thing. Really the more relevant topic here is the finer details of applying it the right way. Now thát is a nice topic. I mean we can all link a bar chart and say the top entry is best.

 

[Deleted member 185158]

The paste I use isnt on the charts in this thread...

Application huh? Thats a different topic from top 2019 thermal paste. You could start a new thread?

AS5 is not a top paste. Thats all anyone really needs to say here. No ifs or ands or buts about it.

 

[Vayra86]

The paste I use isnt on the charts in this thread...

Application huh? Thats a different topic from top 2019 thermal paste. You could start a new thread?

AS5 is not a top paste. Thats all anyone really needs to say here. No ifs or ands or buts about it.

Did people say this, I don't think they did? I think the gist was 'its not relevant enough to switch to something else, and 'this is fine'. And that ties into the application method too. When you're used to the way the substance works, easier application also makes for better mounts. In some cases up to and including a bigger gain than you could get from the best paste.

Enthusiast or not, not everyone is prepared to mount a heatsink half a dozen times to see how to get the best results. But over time, you do gain experience you can use easily in a new build.

But if it satisfies you: I'll say it too. AS5 is not a top paste. Boom!

FYI right now I'm a happy Kryonaut user. And part of the reason is not chart topping performance, but easy application.

 

[Bill_Bright]

Yes, 2c can make all the difference in the world running a comp - I know that firsthand.

Not true! Not unless that 2°C takes you over the thermal thresholds into the "hot" zone that causes instability or throttling. And in that case, it means YOU failed to ensure adequate cooling in the first place. You should never be running where 2 degrees makes a difference. That clearly indicates inadequate case cooling, a bad TIM application, excessive ambient temperature, improper clocks or voltages, mismatched CPU cooler or some other factor due to user failure or incompetence.

Wanting and needing are one in the same when comes to competition.

I already said that. But that is an exception and again, exceptions don't make the rule. You don't "need" to compete.

AS5 is not a top paste. Thats all anyone really needs to say here. No ifs or ands or buts about it.

And nobody in this thread ever said it was! But what you keep ignoring is that it is still a very good TIM and more than adequate for the vast majority or users and TPU readers and members alike.

What you keep ignoring is it is YOUR responsibility to keep the CPU comfortably within its normal operating range, not the TIM's. And comfortably within is not 2°C below thresholds, or even 10°.

 

[Vayra86]

Not true! Not unless that 2°C takes you over the thermal thresholds into the "hot" zone that causes instability or throttling.

There is a caveat to that one. The definition of throttling is being stretched with every new release. It is rapidly being rebadged as 'boost'. In that sense, every C does help, because it gives you performance, however minor it may be.

 

[lexluthermiester]

@Bill_Bright
Let it go man, they're just trying to get a rise out of you for giggles.

 

[Deleted member 185158]

Hey like I said....

Use mustard for all I care lol.

 

[INSTG8R]

FYI right now I'm a happy Kryonaut user. And part of the reason is not chart topping performance, but easy application.

? It’s terrible stuff to apply at least on a bare GPU die...

 

[Vayra86]

? It’s terrible stuff to apply at least on a bare GPU die...

Isn't that conductonaut you are confusing it with?

Hey like I said....

Use mustard for all I care lol.

I reckon youtube has a video on that

 

[Bones]

Understand Bill I mean this with respect, not to get your goat or with any other intent.

Not true! Not unless that 2°C takes you over the thermal thresholds into the "hot" zone that causes instability or throttling.
And in that case, it means YOU failed to ensure adequate cooling in the first place. You should never be running where 2 degrees makes a difference. That clearly indicates inadequate case cooling, a bad TIM application, excessive ambient temperature, improper clocks or voltages, mismatched CPU cooler or some other factor due to user failure or incompetence.

I WAS speaking of Extreme OC'ing, not standard desktop use with this which means not in a case with fans. LN2 pots and all else in use of course.
Like this:

Yes that's one of my setups in action which is the reference to what I was getting at.
Sometimes it does make all the difference between a run that completes or crashes or in some cases will finish but it's bugged all to hell (Invalid result).

I already said that. But that is an exception and again, exceptions don't make the rule. You don't "need" to compete.

This is true - Unless there's a gun pointed at your head there isn't a "Need" to compete - Do it if you want.
It's my stuff and I do.

And nobody in this thread ever said it was! But what you keep ignoring is that it is still a very good TIM and more than adequate for the vast majority or users and TPU readers and members alike.

Show me where in this thread I said AS5 was bad TIM - I never did.

What you keep ignoring is it is YOUR responsibility to keep the CPU comfortably within its normal operating range, not the TIM's. And comfortably within is not 2°C below thresholds, or even 10°.

I haven't ignored that and yes, I agree it is the owners responsibility to maintain the machine or at least get it looked at if a problem develops.

 

[Bill_Bright]

There is a caveat to that one. The definition of throttling is being stretched with every new release. It is rapidly being rebadged as 'boost'. In that sense, every C does help, because it gives you performance, however minor it may be.

I agree they (marketing weenies) are redefining it to include your definition, but that has nothing to do with temperature and everything to do with the "utilization" demands on the processor. When idle, it throttles back to save energy (including battery runtime). Lowering temps is just a welcomed side effect. And when tasked, it throttles back up to full speeds to address the performance needs.

Let it go man, they're just trying to get a rise out of you for giggles.

I wish that were true. But sadly, the comments clearly indicates a lack of understanding and inexperience.

But your advice is still sound and I will, for real this time, step away.

But before I go, for anyone with unopened tubes of AS5 they are going to throw away because it has silver in it , or because it is a whole 5° less efficient that the top and much more expensive performer, please send them my way. I'll put it to good use.

 

[freeagent]

You guys make me want to buy a tube of AS5 just to show you it’s not as bad as you make it out to be.

 

[Deleted member 185158]

@Bill_Bright
Let it go man, they're just trying to get a rise out of you for giggles.

Actually not at all, but just because you say so... it must be!!

I will openly state that AS5 is not a top paste, but never once said it was bad, just like a couple of other fellas whom are only stating a factual point....

AS5 is just simply not a top end paste period end of story.

Just because it "works" for "average" users, doesn't make it bad!

Again. Simple as it can get. AS5 is not a top paste. Not in the top 5 and not in the top 10.

IDC whom has to say what about AS5 being "good enough" that's not the point. It's simply not even arguable the fact it's NOT a top Thermal Interface Material.

@freeagent, really bro?? Nobody said bad paste. It's just not a top paste.

 

[FordGT90Concept]

Again, I present this:

That's AS5 on a stock 6700K running at 100% like it has done almost every day for the last four years. AS5, even fully sintered, may not be the most thermally conductive but it will have the best thermal performance of any TIM over a decade of use (even outperforms solder). So maybe not the best on day 1, maybe pretty close on day 300, but on day 1500 it will still be performing better than the rest.

If you're constantly changing HSFs, don't use AS5. It's expensive and it needs time to sinter. If you put machines together with the intention of never taking them apart again until the motherboard starts on fire, then AS5 is an excellent choice.

All I do is long term computers so all I use is AS5. I don't see a reason to use anything else.


Any test of AS5 needs to take in to account the time it takes to sinter or it doesn't show AS5 at maximum thermal conductivity.

 

[Bones]

You guys make me want to buy a tube of AS5 just to show you it’s not as bad as you make it out to be.

No need to argue, AS5 is good TIM, I've used it before and had no problems with it when I was using it but it's not the best vs others out on the market today.

I don't use it out of personal preference, not by any perceived notion it's dangerous.
Honestly Ceramique is my go-to for daily builds and such, it's cheap and works well enough I don't have any worries with it. For the extreme stuff Ceramique just won't cut it, that's reserved for stuff like Kryonaut and other TIMs designed to work at such extremes.

 

[Raven Rampkin]

That's the thing with some products - overhype. Many products are severly hindered by the overhype.
How many times have you heard: "This video was sponsored by Thermal Grizzly"?
How many times have you heard: "We're using Kryonaut for this video, it is a highly recommended, extreme performance product which you can get with a special *insert Youtuber name here* promo code here yada yada"?
How many times have you heard: "Cooler Master has reached out to us with their new Mastergel"?

Both the Silver and the MX-4 seem to be moreless equally overhyped as well, but with two catches. First, they're overhyped by the community and not the PRO MLG PCMR GRADE REVIEWERZ. Second, the Silver is indeed an older formula AFAIK, and that rubs some people (not me honestly). Otherwise, one would think the Silver has more buzz around it than the MX-4 or vice-versa, but I dunno really. Some people have set the MX-4 as their default option, others have the Silver on default. Some people put "MX-4" at the end of every sentence, others do the same with "Silver 5". Many people ain't touchin' either of these two products.

How many times have you heard: "This video was sponsored by Gelid"?
Me, zero.
Or "We're using MX-2 here", for that matter.
Or "We're using Prolimatech here", for that matter.
Or "We're using NT-H1 here", for that matter, after the initial reviews wore off. Same applies to the NT-H2.
When people use these they're usually quite silent and 'lax about it.
Oh, and Ceramique too. Haven't heard that name too often (yet).

EDIT: Disclaimer, I have never tried the Silver and ain't planning to. Yet, I now see where people are coming from when mentioning that paste. It's actually not about the high or low rating or high or low asking price, IMHO. It's about the hype. Is the Silver, a pretty old formula that neverthless "just works" (c) for many chaps, worth arguing for, shilling for, or hating on, on every single forum board? Nah.
(MX-4, don't think you'll spare the punishment. You ain't far off xD)

 

[Deleted member 185158]

Again, I present this:

That's AS5 on a stock 6700K running at 100% like it has done almost every day for the last four years. AS5, even fully sintered, may not be the most thermally conductive but it will have the best thermal performance of any TIM over a decade of use (even outperforms solder). So maybe not the best on day 1, maybe pretty close on day 300, but on day 1500 it will still be performing better than the rest.

If you're constantly changing HSFs, don't use AS5. It's expensive and it needs time to sinter. If you put machines together with the intention of never taking them apart again until the motherboard starts on fire, then AS5 is an excellent choice.

All I do is long term computers so all I use is AS5. I don't see a reason to use anything else.


Any test of AS5 needs to take in to account the time it takes to sinter or it doesn't show AS5 at maximum thermal conductivity.

Load says 24% at 50c in your pic. My 8700K is at 40c at 5.3ghz with 24% loads lol.

Any how, AS5 is still not a top paste in 2019.

 

[Bill_Bright]

I WAS speaking of Extreme OC'ing

Fine and with respect back, extreme OC'ing is an extreme exception - even among the regular users/posters on this site. And exceptions don't make the rule.
Most users don't overclock, or if they do, it is mild to moderate - perhaps just using the preset options provided by their motherboard BIOS.

Again. Simple as it can get. AS5 is not a top paste. Not in the top 5 and not in the top 10.

This is true. It scored a dismal 37.2°, a full 2.4° warmer than the top Thermal Grizzy Kryonaut's 34.8°C. The 10th down from the top came in at a mind boggling 35.8°C. Clearly more than enough to keep the CPU from totally melting.

The worst, Coolermaster tape, was only able to keep the processor a scorching hot 40.2°C.

Any how, AS5 is still not a top paste in 2019.

Yeah, you've said that over and over and over again - even though still, nobody has ever said otherwise.