[EOL] Arctic MX-5 is here!!Tests incoming! Completed. Now its MX-6 testing time!

[lexluthermiester]

Found these reviews,

I think the general take away is that MX-5 is a solid performer for the price and a great replacement for MX-4.

based on everything I read here, looks like I got excited for nothing. I probably will just sell my mx-5 tube unopened at this point, I have always wanted to use the Noctua NT-H2, bit pricey but eh I'd rather support Noctua as I respect them as a company more than most others.

Why? Keep it, use it, enjoy. Unless you're going to be doing some extreme OCing, the difference in performance will not justify the hassle and expense of getting NT-H2.

 

[Space Lynx]

Found these reviews,

I think the general take away is that MX-5 is a solid performer for the price and a great replacement for MX-4.


Why? Keep it, use it, enjoy. Unless you're going to be doing some extreme OCing, the difference in performance will not justify the hassle and expense of getting NT-H2.

hmm not bad. basically the same as mx-4 overall but with longer shelf life. i guess i will keep it and just use it as my emergency paste as planned. that was main reason i went with it after all... noctua has only a 2 year shelf life i think i read... yikes.

 

[lexluthermiester]

noctua has only a 2 year shelf life i think i read... yikes.

Those estimates can be taken with a big grain of salt. A PC I built for a friend 8 years ago which had a C2Q and a CoolerMaster Geminii had AS-5 which was never changed. It was working perfectly. The only reason for replacement was a much needed upgrade. The AS-5 was a bit hard(as expected) but looked fine. Point being that the estimated "shelf life" is more of a sell-by-date kind of thing and has little bearing on the useful lifespan of the product.

 

[ThrashZone]

Hi,
Depends on how it's stored.

 

[Shrek]

I'm not so sure one really needs a thermal conductivity much above say 5 W/meter K

If the power is 100 W, the area 3cm^2 and the thickness 0.05 mm then the temperature drop will be just over 1 C

I'd rather have a paste that lasts.

 

[Space Lynx]

I'm not so sure one really needs a thermal conductivity much above say 5 W/mK

If the power is 100 W, the area 3cm^2 and the thickness 0.05 mm then the temperature drop will be just over 1 C

I'd rather have a paste that lasts.

what about thermal grizzly conductonaut? I got 20 celsius drop in temps from that on my laptop... desktops don't see that kind of gain but i think they still can get a 5-10 celsius drop from conductnaut depending on the cooler used.

 

[Shrek]

Then something else is going on, as the best improvement over a normal thermal grease is about 1°C

I'm using the equation that defines thermal conductivity

ΔT = P/A Δx/k​

ΔT is temperature drop (in degrees C)
P is the power (in watts)
A is the area (in meters^2)
Δx is the thickness (in meters)
k is the thermal conductivity (in W/meter K)

 

[ThrashZone]

Hi,
Liquid metal is another beast
You'd always see big gains if applied evenly.
Of course lots can go wrong too lol

 

[Shrek]

That stuff scares me

Gallium Induced Failure of a Heat Sink - Bing video

 

[Space Lynx]

That stuff scares me

Gallium Induced Failure of a Heat Sink - Bing video

I'm pretty sure it is fine as long as the only metal it comes into contact with is copper. I think people make the mistake of using non-copper contacts with it... at least this is what I remember when I was putting it on my CPU laptop. which has full copper on both sides.

 

[ThrashZone]

Hi,
Yep nickel or copper only.

 

[Shrek]

I'm happy with SuperLube thermal grease even if I am paying a few °C as a result

Good to 500°F so should last.

 

[Falkentyne]

I'm happy with SuperLube thermal grease even if I am paying a few °C as a result

Good to 500°F so should last.

Hi, I remember you from Jonnyguru forums.
You know your way around power supply hardware very well, but I think your CPU knowledge is a bit lagging.
That Super Lube gel is terrible. Other people have tested it. There was a thread about it long ago and it was atrocious. Compared to Kryonaut, on something like a 9900k, it's like more than 10C worse. And it's only 0.55 w/mk. There was an article (maybe around 2015) which showed Super Lube 10C worse than a typical decent thermal paste (pictures dont work anymore but the review was linked on tomshardware to here: https://overclockersonline.net/?p=50117671 ) And I remember this being discussed on another main site once---terrible.

This type of gel is ONLY suitable for heavy machinery or electronics with low power density. Absolutely NOT for any modern CPU nor video card.

Also, while your w/mk formula you gave is accurate, it doesn't take into account other factors that go into thermal paste material.

And you can clearly see from this chart right here that your w/mk theory falls right on its face in real life in real world usage.

TechnologyGuide

Thank you for visiting the TechnologyGuide network. Unfortunately, these forums are no longer active. We extend a heartfelt thank you to the entire community for their steadfast support—it is really you, our readers, that drove

forum.notebookreview.com

Just to put this in perspective, IC Diamond is 4.5 w/mk. Yet look how well it does. But you don't want to use it on a CPU IHS. There it's w/mk makes it fall behind the main paste leaders (Kryo Extreme, TFX, MX-5, Kryonaut, Gel Maker Nano, etc).


Alphacool "Subzero", aka " Phobya Nanogrease Extreme" (it's the same crap) is 16.5 w/mk.
It's almost as thick as Thermalright TFX, yet it does much worse. Terrible.

And I've tested Nanogrease Extreme on a r9 290X vs Kryonaut. The nanogrease extreme always performed worse (like 2C or something).
On my r9 290X, Arctic MX-5 was only about 1C worse than Kryonaut. Thermalright TFX is better on my 10900k after curing a week (compared to Kryonaut) and otherwise trades blows with Kryonaut Extreme.

 

[Shrek]

I concede that the SuperLube offering is far from ideal; that is why I said I felt that going beyond 5 W/ meter K was probably not worth it (for paste);
and you are right, I assumed 1/20th of a millimeter thickness, while thick pastes may make this difficult.

However, my Core 2 Quad only reaches around 70°C when stressed by CPU-Z and idles around 40°C (on SuperLube)

The problem with all the reviews: I want to know which pastes will last indefinitely and is the reason I went with SuperLube;
I am wondering about carbon thermal pads as they should last, but the 35 W/meter-Kelvin is deceptive as the thickness is more than for thermal paste.

• So let's calculate for 0.2 mm carbon pads

P = 100 W​

A = 0.03^2 m^2​

k = 35 W/meter Kelvin​

Δx = 0.2 10^-3 m​

to yield less than 1°C drop; so maybe this is the direction I should take.

I think Indium foil can also be used as a thermal pad (82 W/meter-Kelvin), as thin as 0.05 mm on ebay

• So let's calculate for an 0.05 mm Indium pad

P = 100 W​

A = 0.03^2 m^2​

k = 82 W/meter Kelvin​

Δx = 0.05 10^-3 m​

yielding less than a 1/10th of a degree; that would be hard to beat so I'd be tempted to go with the thicker 0.1mm offering;
but I don't know if Indium is soft enough to mold to the surfaces it interfaces (it is softer than lead). Then again it might be best to avoid an electrical conductor.

Concerning diamond containing paste, I have read that it can scratch what it is in contact with.

On a side issue:
It bugs me to see W/mK since I read mK as 'milli-Kelvin' and not 'meter Kelvin'

Any idea if JonnyGuru will ever come back?


Now you have made me wonder if the paste/metal interface suffers a significant temperature drop...

 

[8BitZ80]

Then something else is going on, as the best improvement over a normal thermal grease is about 1°C

I'm using the equation that defines thermal conductivity

ΔT = P/A Δx/k​

ΔT is temperature drop (in degrees C)
P is the power (in watts)
A is the area (in meters^2)
Δx is the thickness (in meters)
k is the thermal conductivity (in W/meter K)

If only it were that simple. Thermal conductivity is just one variable in thermal transfer across multiple boundaries, and on top of THAT, there is no standardized method for calculating thermal conductivity. Those values have such little meaning that I don't even read them.

 

[lexluthermiester]

I concede that the SuperLube offering is far from ideal

Regardless, does it do the job? If it does the job you need and does it well then it is suited for your case scenario. Put another way, if it works for you then carry on and quit giving consideration to nay-sayers...

 

[8BitZ80]

Regardless, does it do the job? If it does the job you need and does it well then it is suited for your case scenario. Put another way, if it works for you then carry on and quit giving consideration to nay-sayers...

Brand loyalty is really strong when it comes to thermal pastes. Even the best thermal paste in the world won't be noticed by many people if it's from an obscure brand.

There's a bit of psychology involved too, because generally a higher priced product is perceived by consumers to be higher quality. I recall a story my relatives told me about how they had bee hives and couldn't sell any of the honey they produced, and it only started getting popular after they doubled the price. Consumers are just hard-wired to assume that a cheap price by default means rubbish quality. I see exactly the same mentality when I recommend the GD900 bulk thermal paste which is ridiculously cheap compared to retail pastes, I can present all the benchmarks in the world to prove that it works really effectively but people will still assume that it's a rubbish product and they should avoid it. That's just how consumers are.

 

[lexluthermiester]

Brand loyalty is really strong when it comes to thermal pastes. Even the best thermal paste in the world won't be noticed by many people if it's from an obscure brand.

There's a bit of psychology involved too, because generally a higher priced product is perceived by consumers to be higher quality. I recall a story my relatives told me about how they had bee hives and couldn't sell any of the honey they produced, and it only started getting popular after they doubled the price. Consumers are just hard-wired to assume that a cheap price by default means rubbish quality. I see exactly the same mentality when I recommend the GD900 bulk thermal paste which is ridiculously cheap compared to retail pastes, I can present all the benchmarks in the world to prove that it works really effectively but people will still assume that it's a rubbish product and they should avoid it. That's just how consumers are.

My experience is that people are results driven and when something works for them they tend to stay with it, especially when that something works well. Brand loyalty only factors in when a particular brand performs well consistently.

 

[joemama]

P = 100 WA = 0.3^2 m^2k = 35 W/meter KelvinΔx = 0.2 10^-3 m

You've got your area wrong, 0.3^2 square meters is 30x30cm

 

[Shrek]

You are right; it was a typo, the calculations were done with 0.03^2 so the results are the same (fixed the original); much appreciated.

My desire is longevity and none of the reviews cover this, for obvious reason.

Regardless, does it do the job? If it does the job you need and does it well then it is suited for your case scenario. Put another way, if it works for you then carry on and quit giving consideration to nay-sayers...

Ah, but it is very important to listen to the critics, even if they are wrong; the faster I find my own error (if there is one) the faster I find the solution and the lesser of a fool I seem ;-)


Now we have GD900-1 which is a lot more expensive than GD900; anyone tried this?

• GD900 (4.8 W/m.K)
• GD900-1 (6 W/m.K)

 

[lexluthermiester]

Ah, but it is very important to listen to the critics, even if they are wrong; the faster I find my own error

Perhaps. But if the compound you are using is working for you, and working well, there is no error. Maybe room for improvement, but certainly not an error.

the faster I find the solution and the lesser of a fool I seem ;-)

Why do you care if anyone thinks you a fool? Use what works for you.

 

[Shrek]

Because I have seen people who don't care and feel they are always right... it is not a pretty thing to see; so I try to be right, but am delighted to be shown wrong.

The video card I just brought back to life with a reflow reaches 90°C using FurMark so I am wondering if I need something better than SuperLube. In reality FurMark is perhaps unrealistic but it was a great way to stress test the card after the reflow repair.

 

[lexluthermiester]

The video card I just brought back to life with a reflow reaches 90°C using FurMark so I am wondering if I need something better than SuperLube.

Furmark is a known card-killer. I've seen this first hand. It is an extreme stress testing utility. 90C is actually not bad(depending on the card). There's nothing wrong with the TIM you are using. Better cooling is needed. Word of caution though, you should avoid using Furmark for anything more than a few minutes(read 3 or 4) at a time or you risk permanent damage to the GPU tested.

 

[Shrek]

Oops... I let it run for several hours to be sure the reflow was good.

 

[lexluthermiester]

Oops... I let it run for several hours to be sure the reflow was good.

Oh, wow... Don't do that again.. One can get away with 5 or 6 minutes but several hours is asking for trouble. And for a GPU you just reflowed/repaired, no need pushing your luck. Although, if it's still stable and running after that level of brutal torture, you can feel confident about your repair work!