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

[fluxc0d3r]

SYY should give you similar results as TFX, though TFX is still slightly better. Both are top-tier thermal grease. I don't think SYY and Zezzio are related, because the conductivity ratings are different.

Issue with SYY is that it is really thick, and not easy to apply. The Zezzio was quite easy to apply.

The ONLY thermal grease that I WANT is super hard to find is the Cougar TC-5888, which is made by DOW. Just released and seen on JD.com. It's highly recommended by Asian overclockers.

Here's a link for the Cougar TC-5888: https://item.jd.com/10029456260263.html

I've seen a lot of new Youtube videos where Chinese system builders are using this stuff and wondering where to find it. Guess it's only available in Asian markets for now.

 

[ryzenmaster]

Thanks this is great !

Any suggestions how better apply SYY on each die 5900hx and the RX6800M?

Put SYY in hot boil water cup first before apply to make it easier to apply?

Photos attached from previous lab service repasting day...gelid gc extreme used here.

BTW crazy noob question: anyone tried mix different pastes ?

 

[fluxc0d3r]

 

[Falkentyne]

SYY should give you similar results as TFX, though TFX is still slightly better. Both are top-tier thermal grease. I don't think SYY and Zezzio are related, because the conductivity ratings are different.

Issue with SYY is that it is really thick, and not easy to apply. The Zezzio was quite easy to apply.

The ONLY thermal grease that I WANT is super hard to find is the Cougar TC-5888, which is made by DOW. Just released and seen on JD.com. It's highly recommended by Asian overclockers.

Here's a link for the Cougar TC-5888: https://item.jd.com/10029456260263.html

I've seen a lot of new Youtube videos where Chinese system builders are using this stuff and wondering where to find it. Guess it's only available in Asian markets for now.

Conductivity ratings are very often hogwash.
The Zezzio thermal pads are rated as 14.8 w/mk but are the exact same temp performance as Gelid Extreme's 12 w/mk pads, on a 3080 FE video card. They were tested both on the exact same card.

I also as I mentioned earlier, have several tubes of SYY-157 and I bought the Zezzio "14.3 w/mk" thermal paste as soon as it appeared on Amazon. It's the exact same as SYY-157, and the temps are exactly 0C different. Also I don't know why I'm the only one who caught it, but Zezzio simply copied TFX's back label *exactly*. Their wording on the back of their packaging is 100% identical to what Thermalright puts on their packaging. Including the "HARMLESS: YES" part. That's why they have it as the same w/mk as TFX. But TFX slightly outperforms both SYY-157 and Zezzio on laptops and GPU's, although they are almost the same on desktop CPU's. It also has the exact same physical properties, when you check by doing a fingerprint spread test. (Compare that to TFX, you will notice that TFX leaves a 'thicker' sheen on your finger, while SYY/Zezzio leaves a thinner sheen. FuzeIce Plus / Alseye T9+ Platinum (which also seem to be identical) leave a much wetter sheen, and of course they are the easiest to spread. All of those pastes are nano-carbon, rather than traditional silicone pastes.

 

[lexluthermiester]

View attachment 227548

I don't know where this graph came from, but it does not agree with information found elsewhere about the performance of TFX. This graph is rubbish and nonsense until properly cited. That said...

Any suggestions how better apply SYY on each die 5900hx and the RX6800M?

Put SYY in hot boil water cup first before apply to make it easier to apply?

...this should work and SYY should provide reasonable performance for your system. Keep in mind that thermal paste is not a magic wonder solution that will solve all your cooling problems. It's just a tool to help make thermal transfer more efficient. The limiting factor for your desired temps is the cooling system employed in that laptop. If that heatsink/fan setup can't drop your temps to levels you desire, you're only realistic solution is to downclock the CPU/GPU so that they don't create as much heat.

 

[ryzenmaster]

I don't know where this graph came from, but it does not agree with information found elsewhere about the performance of TFX. This graph is rubbish and nonsense until properly cited. That said...


...this should work and SYY should provide reasonable performance for your system. Keep in mind that thermal paste is not a magic wonder solution that will solve all your cooling problems. It's just a tool to help make thermal transfer more efficient. The limiting factor for your desired temps is the cooling system employed in that laptop. If that heatsink/fan setup can't drop your temps to levels you desire, you're only realistic solution is to downclock the CPU/GPU so that they don't create as much heat.

Ok so at least now I have even more hope that SYY will drop some Celsius versus gelid gc extreme

Any particular apply method that is better in my case ?

 

[lexluthermiester]

Any particular apply method that is better in my case ?

If SYY is thick, warming it up like you suggested above might be helpful. Just make a smooth thin layer on the CPU/GPU dies and put the heatsink on. The mounting pressure will take care of the rest. You're good after that.

 

[ryzenmaster]

If SYY is thick, warming it up like you suggested above might be helpful. Just make a smooth thin layer on the CPU/GPU dies and put the heatsink on. The mounting pressure will take care of the rest. You're good after that.

Thin layer on cooling chamber side too?
Hello about add some dots ?

 

[lexluthermiester]

Thin layer on cooling chamber side too?

Can't hurt.

Hello about add some dots ?

I don't do the dots thing. I spread a thin layer to guarantee even and complete coverage. Air pockets are never a thing as mounting pressure from the heatsink mechanism will push any air out along with the excess thermal paste. If you look back to my posts here, you will see what I mean. When I say "Thin Layer" I really do mean thin.

 

[ryzenmaster]

Can't hurt.

I don't do the dots thing. I spread a thin layer to guarantee even and complete coverage. Air pockets are never a thing as mounting pressure from the heatsink mechanism will push any air out along with the excess thermal paste. If you look back to my posts here, you will see what I mean. When I say "Thin Layer" I really do mean thin.

You sure this approach of so thin layer without a dot works for laptops? as laptop has not so high pressure mounting, wouldn't an extra dot in middle assure better filler?

 

[lexluthermiester]

You sure this approach of so thin layer without a dot works for laptops? as laptop has not so high pressure mounting, wouldn't an extra dot in middle assure better filler?

I'm sure.

 

[freeagent]

You want it thin but not translucent. I draw a line down the side and pull it over to the other side with a spatula. There is such a thing as too much, trust me.

 

[ryzenmaster]

You want it thin but not translucent. I draw a line down the side and pull it over to the other side with a spatula. There is such a thing as too much, trust me.

Any photos or illustrations?

 

[lexluthermiester]

You want it thin but not translucent.

This.

Any photos or illustrations?

See my above post.
https://www.techpowerup.com/forums/posts/4534033
These photos should give you an idea.

 

[freeagent]

Any photos or illustrations?

I haven't taken any, sorry

Mostly because I didn't think anyone would care

 

[ryzenmaster]

Ok folks, testing is done.

Let's state the particulars.

The system being used is a Dell Precision T3500 with an Intel Xeon W3680 using the stock cooler and a steady state 3700RPM 92mm fan. The testing software was Prime95. Temp monitoring software is HWInfo. The thermal interface materials being tested are as follows:
Dragon Star DRG102, compound very similar to Arctic Silver 5
Automotive Copper thread grease
Noctua NT-H1
Arctic Cooling MX-5

The tests were conducted as follows:
3 runs of 2 stages of Prime95 calculations which run for a little over 11 minutes per each TIM.

Here is how each was applied. I did not take photos of the Noctua NT-H1 as I forgot to. It didn't look any different than the DRG102.

DRG102
View attachment 202525View attachment 202526

Copper Grease
View attachment 202527View attachment 202528

MX-5
View attachment 202529View attachment 202530

As you can see, I spread the TIM out in very thin layers. The DRG102 and Copper Grease were the easiest to spread evenly. NT-H1 was a little more difficult but only a bit. MX-5 was a bit of challenge but with a few minutes of work it spread out fine. It is as "sticky" as everyone claims of course, however this is a minor thing.

Here are the test results.

Dragon Star DRG102
View attachment 202519
This performance ranks third best in the tests.

Automotive Cooper thread grease
View attachment 202520
This performance ranks fourth in the tests.

Noctua NT-H1
View attachment 202521
This performance was second best in the tests.

Arctic Cooling MX-5
View attachment 202522
This performance was the best of the tests.

My conclusion is that MX-5 is a very impressive TIM. It bested NT-H1 by several degrees C in both idle and full load tests. It also bested the DRG102 handily. The copper grease proved inferior to the rest, even though the temps were within reason as expected given that the compound was used in this same system for nearly a year and performed well. NT-H1 is widely regarded as a premium TIM and has shown itself to be an excellent performer. For MX-5 to perform better than such shows it to be, at the very least, competitive with premium TIM's and at most superior to many of the of the top shelf brand TIMs.

The screenshots are here for you all to draw your own conclusions though.

IMPORTANT EDIT;
For the record, this testing was done as much for fun as it was out of scientific, professional and personal curiosity. The testing methodologies used are based on known valid methods but are NOT all inclusive. I'm NOT going to test every TIM on the market, nor am I going to test on a multitude of platforms. These tests are focused on the topic of discussion here, Arctic Cooling MX-5. I deliberately chose to use a TIM that is a budget known good performer, a product not specifically designed for the purpose at hand(for giggles and to be an odd-ball) and a TIM that is considered to be a premium product for testing against MX-5.

That said, if you have a problem or complaint about the testing methods used or the fact that the products tested were limited, feel free to buy your own products to test, spend your own time conducting said testing and then post your own results. Whining about what I did or didn't do that somehow doesn't meet with your satisfaction will only earn you an unpleasant reaction, which will likely include open and public mocking.

Only warning.

But should not paste cover whole die?

 

[lexluthermiester]

But should not paste cover whole die?

Of course. Keep in mind, you're working on a laptop with a bare die whereas I was working with a desktop CPU in that testing run. However, I always use the same spreading method across laptop CPU/GPU die that I use on CPU's with an IHS.

 

[Shrek]

Here is how each was applied. I did not take photos of the Noctua NT-H1 as I forgot to. It didn't look any different than the DRG102.

DRG102
View attachment 202525View attachment 202526

Copper Grease
View attachment 202527View attachment 202528

MX-5
View attachment 202529View attachment 202530

Interesting, I never felt the need to cover both surfaces; I just cover the chip area as I am concerned more on the heatsink may just attract dust.

Whining about what I did or didn't do that somehow doesn't meet with your satisfaction will only earn you an unpleasant reaction, which will likely include open and public mocking.

Only warning.

I is ready

I may have asked before, but is the copper grease electrically conductive?

 

[Metroid]

After all, I have to say that MX5 was disappointing. I prefer MX4 more than this.

I was about to buy the mx-5 and I have a mx-4, after I have seen few reviews, I decided not to buy the mx-5 anymore. They are pretty close and that is within the margin of error, so no point if you already have a big tube of mx-4(my case), that will last for me at least 3 years down the line.

 

[Ruru]

I was about to buy the mx-5 and I have a mx-4, after I have seen few reviews, I decided not to buy the mx-5 anymore. They are pretty close and that is within the margin of error, so no point if you already have a big tube of mx-4(my case), that will last for me at least 3 years down the line.

Exactly. I bought a 8g tube of MX4 the last time when I bought paste, and I'll stick with this also in the future.

The construction of MX5 was also annoying. MX4 behaves like traditional paste.

 

[lexluthermiester]

Interesting, I never felt the need to cover both surfaces;

My experience has been that it just works better.

I just cover the chip area as I am concerned more on the heatsink may just attract dust.

How? The surfaces will be joined together and in direct contact. How would dust get in to that space at all and in a way that is different from applying TIM only to the CPU?

I was about to buy the mx-5 and I have a mx-4, after I have seen few reviews, I decided not to buy the mx-5 anymore. They are pretty close and that is within the margin of error, so no point if you already have a big tube of mx-4(my case), that will last for me at least 3 years down the line.

Gentlemen, @W1zzard uses MX-5 as a default for testing as of late. He can afford & use whatever he wishes and he defaults to MX-5. You can call my testing and conclusions into question all you wish, but you then have to question all of W1zzard's recent testing as well.

Give that a moment of thought.

 

[Shrek]

How? The surfaces will be joined together and in direct contact. How would dust get in to that space at all and in a way that is different from applying TIM only to the CPU?

I imagine the area covered on the heatsink might be larger than needed; this overhang might attract some dust, even if it does not affect the cooling.

Maybe one fine day you will add GD900 to your test list.

Might I ask you throw the probes of your digital multi-meter into the copper grease to tell us if it is electrically conductive or not.

 

[Ruru]

Come continue your fight on my thread guys

The Official Thermal Interface Material thread

No need for another single thread, let's use this for questions and suggestions. :)

www.techpowerup.com

 

[Shrek]

I've been suspended from a thread before, so while I would love to; I would not dare...

 

[Ruru]

I've been suspended from a thread before, so while I would love to; I would not dare...

Well, just keep it cool this time when arguing about pastes etc (pun intended)