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

[freeagent]

I ran it once on my 980 when I first got it.. it tripled its power consumption at the wall and I will never run it again

The bios switch was not in its stock position and I think it was getting full power.. I had an 850w PSU and my meter was showing 750w with an overclocked 3770K..

Edit:

I may have had the CPU loaded to just to see what the worst case scenario would be.. because I was silly and did things like that.

 

[Shrek]

Imagine a virus whose task it was to destroy the video card... a card should be able to take full load.

 

[freeagent]

That's what I thought too. But the big numbers scared me

To me it was not natural lol.. but yet I run Linpack quite often and that too is not natural..

 

[Shrek]

This pump out effect is concerning

(1) Der8auer Deep-Dives on Thermal Paste: Misconceptions, Curing, & More | LTX 2019 - Bing video

Makes me wonder about Diaper cream that is 40% zinc oxide because it will dry out and no more pump out.

Yes, I'm just kidding; but it might be why manufacturers use thermal compound that sets.

 

[Falkentyne]

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.


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)

Longevity?
Buy Thermalright TFX. Very long lasting paste. Very durable with good contact pressure being used. Just make sure the contact pressure is firm and even and proper.

IC Diamond is probably the most durable but that's up to debate.
I heard the extremely old Arctic Ceramique lasts forever though. Ceramique 1, not 2.

 

[8BitZ80]

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.


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)

Thanks for the tip, you better believe that I'm going to (and already did) buy some for testing. I didn't know about GD900-1 until now. GD900 is already on par with MX-4 and there's only a few random videos in other languages claiming a 1-2C improvement with GD900-1. I'll find out if that's true in due course.

 

[Shrek]

Thanks for the tip, you better believe that I'm going to (and already did) buy some for testing. I didn't know about GD900-1 until now. GD900 is already on par with MX-4 and there's only a few random videos in other languages claiming a 1-2C improvement with GD900-1. I'll find out if that's true in due course.

Of course I went and messed everything up by bringing in yet another factor, namely 'the pump effect'

That said, I very much appreciate your testing candidates; I'm tempted to stock up on the old GD900

I'm still trying to find out if there is a significant temperature drop at the paste/metal interface... which would invalidate the straight thermal conductivity calculations.

Cost is a big issue for me; I got my Core 2 Quad CPU for $25 and Thermalright TFX is $14 for just 2g and even so is rated for just 8 years... my computer is already 12 years old and still going strong. 85g of SuperLube was $7 and should last longer than 8 years.

 

[lexluthermiester]

IC Diamond is probably the most durable but that's up to debate.

Sorry, I will not use IC Diamond for any reason and can't in good faith recommend anyone try it either. A few years ago they made a TIM that was doing physical damage to IHS surfaces and bare dies that either came stock that way or were delidded CPU's. There was a big scandal about it and the company's reaction, response and behavior was enough to make me want to avoid them. There was a big bit of drama here in TPU's forums over the matter.

I heard the extremely old Arctic Ceramique lasts forever though. Ceramique 1

True! I still have a tube of it that still works fine. I don't use it as it's performance is lesser than newer TIM's, but it's not useless and remains viable for a very long time.

 

[thesmokingman]

Imagine a virus whose task it was to destroy the video card... a card should be able to take full load.

Do you drive your car flat out for 500 mile testing to see if it should just work? GPU like any device or product is made for the expected environment. They aren't equipped with coolers that can handle something like furmark for hours on end.

 

[Shrek]

Exactly why I mentioned a virus... I was reluctant to voice this concern as it might give someone ideas.

Some programs offload calculations to the GPU, mining comes to mind; so yes, I can see scenarios where the GPU needs to be able to carry a full load for extended periods of time.

 

[8BitZ80]

Of course I went and messed everything up by bringing in yet another factor, namely 'the pump effect'

That said, I very much appreciate your testing candidates; I'm tempted to stock up on the old GD900

I'm still trying to find out if there is a significant temperature drop at the paste/metal interface... which would invalidate the straight thermal conductivity calculations.

Cost is a big issue for me; I got my Core 2 Quad CPU for $25 and Thermalright TFX is $14 for just 2g and even so is rated for just 8 years... my computer is already 12 years old and still going strong. 85g of SuperLube was $7 and should last longer than 8 years.

I wouldn't get too hung up on details, opinions and rumors. There's more than enough conflicting information in thermal paste land to drive you clinically insane. The reality is that there's plenty of good products out there that will do the job. I haven't had much experience with the longevity of thermal pastes, but the only ones I've noticed that have gone bad were the stock Intel heatsink pastes that had expired after five years.

Why not give MX-2 or MX-4 a go? They perform good and will last a long time in your main rig, and then keep the Superlube stuff as backup or for other projects.

 

[Shrek]

Might just do that, but not till you have reported on GD900

Nobody seems to be mentioning the Halnziye HY series
Shenzhen Halnziye Electronics Co., Ltd._Others

HY410 1.42 W/K.m
HY420 1.63
HY430 1.84
HY450 2.15

HY510 1.93
HY520 2.3
HY530 2.5
HY550 2.7

HY610 3.05
HY620 3.34
HY630 3.57
HY650 3.76

HY710 3.17
HY720 3.46
HY730 3.65
HY750

HY810A 4.15
HY810B 3.78
HY810 nano 4.63

HY880 5.15
HY881
HY882 6.0
HY883 6.5

why so many? and I have not listed them all


Concerning the clinically insane part... I fear it is already too late...

 

[8BitZ80]

Might just do that, but not till you have reported on GD900

I'm still working on my methodology and verifying that the readings are consistent. But from what I've seen MX-4 and GD900 perform basically the same, maybe with a tiny 1 degree advantage to GD900. Those are unofficial results so the magnitude of difference might change.

 

[Space Lynx]

from what i can gather on youtube it looks like the best paste for desktop PC is kyronaut extreme... i just went to go buy some, and it costs $30 for 2 gram tube. RIP

i'll pass

 

[nguyen]

from what i can gather on youtube it looks like the best paste for desktop PC is kyronaut extreme... i just went to go buy some, and it costs $30 for 2 gram tube. RIP

i'll pass

Kingpin KPx is the best, and it cost 30usd for 10g, good stuff.

 

[Space Lynx]

Kingpin KPx is the best, and it cost 30usd for 10g, good stuff.

link? I can't find it at that price.

 

[Shrek]

I'm still working on my methodology and verifying that the readings are consistent. But from what I've seen MX-4 and GD900 perform basically the same, maybe with a tiny 1 degree advantage to GD900. Those are unofficial results so the magnitude of difference might change.

Don't forget to include diaper cream but only the good stuff (40% zinc oxide) and vaseline.

Price is a big factor
"An engineer can do for a dollar what any fool can do for two."
Arthur Mellen Wellington

OK, I couldn't take it no more... ordered me four 30g pots of GD900 ($3.51 a pot with shipping)


Now I find

GD32 1.0 W/K.m
GD33 1.0
GD66 1.05
GD100 1.094
GD220 1.114
GD280 1.159
GD380 2.2
GD460 2.0
GD450 2.0
GD900 4.8
GD900-1 6

 

[8BitZ80]

I've run tests on a few different pastes and these were my results after adjusting for 25C ambient temperature:

Thermal Grizzly Kryonaut ($12.00/g) = 52.2C
Arctic MX-4 ($2.20/g) = 52.9C
Kingpin KPX ($6.66/g) = 53.0C
GD900 ($0.15/g) = 54.0C

Mind you that this is running Furmark for 2 minute bursts on an R9 270X (180W TDP with a 212mm2 die area) with fan speed at 100% and I'd estimate that the results are accurate down to around +/- 0.1C. Even after multiple runs, my readings were within 0.1C of each other. I also added the prices I paid for them.

It's pretty obvious that both the budget options (MX-4 and GD900) put on a very good showing. Kryonaut did come out on top so I'll give it credit for that, but the micro-sized pipette of paste is a joke. I thought KPX was okay, but the cost was almost $10/g once you include postage and the temperatures were average.

I was thinking about doing even more testing, but the differences are so minor and it's such a yawn-fest that I reckon I've already seen enough. I will definitely be doing an MX-4 versus MX-5 comparison once my MX-5 finally gets here though, that's a test that definitely needs to happen.

 

[Shrek]

Would be nice to include say Vaseline in the tests.

 

[Space Lynx]

I've run tests on a few different pastes and these were my results after adjusting for 25C ambient temperature:

Thermal Grizzly Kryonaut ($12.00/g) = 52.2C
Arctic MX-4 ($2.20/g) = 52.9C
Kingpin KPX ($6.66/g) = 53.0C
GD900 ($0.15/g) = 54.0C

Mind you that this is running Furmark for 2 minute bursts on an R9 270X (180W TDP with a 212mm2 die area) with fan speed at 100% and I'd estimate that the results are accurate down to around +/- 0.1C. Even after multiple runs, my readings were within 0.1C of each other. I also added the prices I paid for them.

It's pretty obvious that both the budget options (MX-4 and GD900) put on a very good showing. Kryonaut did come out on top so I'll give it credit for that, but the micro-sized pipette of paste is a joke. I thought KPX was okay, but the cost was almost $10/g once you include postage and the temperatures were average.

I was thinking about doing even more testing, but the differences are so minor and it's such a yawn-fest that I reckon I've already seen enough. I will definitely be doing an MX-4 versus MX-5 comparison once my MX-5 finally gets here though, that's a test that definitely needs to happen.

keep in mind that is the old kyronaut, kryonaut extreme probably wins another celsius point off.

 

[nguyen]

extreme overclocker Luumi tested some thermal paste from China which I highly recommend (Thermagic ZF-EX, if you can find them cheap enough, only 1C behind Kryonaut Extreme)

In Vietnam I can buy a 2g tube of this Thermagic ZF-EX for 6usd, this paste is highly viscous that ensure it's longevity, especially on GPU and Laptop.

 

[8BitZ80]

If everyone reported relative to an agreed upon 'base standard' things might be comparable, i.e. independent of area, power.

Perhaps x times better than say SuperLube heatsink compound, or Vaseline

Arctic Silver Alumina states something along those lines:

Lower full load core temperatures
2 to 10 degrees centigrade lower CPU full load core temperatures than standard thermal compounds or thermal pads.
When measured with a calibrated thermal diode imbedded in the CPU core.

My piece of string is longer than a standard piece of string. Confirmed!

keep in mind that is the old kyronaut, kryonaut extreme probably wins another celsius point off.

Ah, that's probably what they mean when people are discussing old vs new Kryonaut. I wouldn't doubt it's pretty good. Shit's expensive though, it's like $20.00/g and that's way beyond my budget.

UPDATE: I also tested MX-2 and it's performance is totally identical to MX-4. Although the MX-2 is a bit harder to spread but it also costs less ($2.00/g).

 

[Shrek]

extreme overclocker Luumi tested some thermal paste from China which I highly recommend (Thermagic ZF-EX, if you can find them cheap enough, only 1C behind Kryonaut Extreme)

View attachment 198730

For me a couple of degrees is not very significant, so only one paste is no good in my opinion.

 

[nguyen]

For me a couple of degrees is not very significant, so only one paste is no good in my opinion.

High quality thermal paste will only make discernable difference in worst case scenario, like GPU with poor stock cooler and laptop, where operating temperature is normally in the 80C range. Using good and highly viscous TIM in this case can shave off 5-15C easily.
But yeah on CPU it really doesn't make any difference though, unless you are deliding older Intel CPU.

 

[Shrek]

I have such a video card and GD900 on the way... 5-15C is definitely worth it