Arctic Silver 5 (aka Old Gray Beard) vs Arctic MX6 (aka Young and Hung) - The Big Show!

[DWD1961]

I've always been a fan of Arctic Silver 5 since it came out about, oh, 100 years ago. Yeah, I'm getting old. I've tried many of them and the Arctic Silver 5 always performed nearly as well if not better than the latest 'space age' pastes. What I like about Arctic Silver is that after storing a computer in a garage for 6 years, the Arctic Silver 5 was still good. This shit lasts virtually forever.

Anyway, the newest 'space age' paste drew me in. I'm going to test Arctic Silver 5 with the newest Arctic miracle paste, the. . .wait for it --

ARCTIC MX6! (As you all probably know, Arctic is not the same company as Arctic Silver. I didn't know that until I started researching it.)

Anyway, I actually contacted Arctic Silver and asked them where I could purchase authentic Arctic Silver 5. I bought some from Platinum Micro,, which is an authorized reseller. The Arctic Silver rep said Arctic Silver source it directly from Arctic Silver, so it's the good.
I also bought some MX6 from Arctic itself, so again, it's the good.
I'm going to test it with my rad fan at 100% in idle and under P95 stress. Then I'm going to test it with the fan at 50% at idle and under stress. It's the best i can do. I know it would be better to have a system where the CPU is bouncing off its thermal limits, but I have what I have. I'll test it using a laser thermometer aimed at the table the computer is on several feet away from the computer. I did that a couple nights ago and the table was the same temperature as my home digital thermostat reported. Again, that's the best I can do.

Anyway, I'll post the results above. If anyone has any test parameters I should use, please post!

___________________________RESULTS__________________________

​

Not too surprising, AC5 did really well, although the MX-6 is better. However, it was about a 1C difference between the two at idle and max heat.
CASE: Thermaltake The Tower 100 (One fan, which was the radiator fan)
Cooling: Enermax 120 AIO with the included Enermax 120 SQu 2200RPM fan. No other case fans.
CPU: Late Gen 3600 fresh from AMD (12/2022)
Parameters

• I took a ambient reading using a laser probe about 2 feet from the computer on the table. It registered 21.8. I took several reading throughout the tests and they varied by 0.2 C.​
• I used HWiNFO64for the numbers. I used max and min heat readings.​
• Prime 95 on max heat setting​
• After each max heat run, I let the PC idle until the lowest temp was reached. I shut down all foreground programs, such as browsers, streaming programs, and file explorers.​
• Paste application was the same for both pastes.​
• MAX Heat was tested with fan and pump at max​
• Idle Temps were taken with fan and pump at 100% and then at what is silent for me, which is fan 50% 1200RPM and pump 66% 2200RPM​

PASTE TEMP DIFFERENTIAL
There was about a 1.5C difference in favor of the MX-6 paste.

PROBLEMS WITH TEST
Test configuration never really gets hot and either paste may perform worse of better as temps increase beyond what I tested. Also, the difference in this test did not show any statistically convincing difference, even though the max and min temps both showed about a 1C to 1.5C difference in favor of MX-6.

SHOWN AS TESTED

TEMP PROBE

PASTE APPLICATION

MAX TEMP

MINIMUM IDLE TEMP FAN/PUMP 100%

MINIMUM IDLE TEMP FAN 50%, PUMP 66%

 

[WhoBeDaPlaya]

Still have several sponsored tubes of AS5 from my reviewing days (early 2000s - BlueSmoke.net, now defunct)

 

[pendragon1]

how about some temps from something like hwinfo instead of a thermometer aimed at the table?!

 

[DWD1961]

how about some temps from something like hwinfo instead of a thermometer aimed at the table?!

LOL, do you really think I'm taking CPU temps with an external laser probe pointed in the general area as the computer? I guess I wasn't too clear. I'm taking the ambient temp with an laser probe, and then I'll use HWiNFO64 for CPU temps.

 

[pendragon1]

LOL, do you really think I'm taking CPU temps with an external laser probe pointed in the general area as the computer? I guess I wasn't too clear. I'm taking the ambient temp with an laser probe, and then I'll use HWiNFO64 for CPU temps.

lol ok but you never know around here...

 

[xDiVolatilX]

I just went with the Noctua NTH1 paste not because I have a Noctua cooler, It's an EK Quantum Vector 2 water block, but because it had nothing but 5 star reviews thousands of them. I'll admit, this paste is awesome. Temps are very good. It's nice and think consistency. Easy to work with. All around perfect performance. Also have arctic 5 used it for years also good lol

 

[Spirit_Retro]

lol ok but you never know around here...

I usually just tape an old glass rectal thermometer to the back plate.

You people have lost your hacker roots.

 

[DWD1961]

lol ok but you never know around here...

Well, it was funny at least.

 

[DWD1961]

I just went with the Noctua NTH1 paste not because I have a Noctua cooler, It's an EK Quantum Vector 2 water block, but because it had nothing but 5 star reviews thousands of them. I'll admit, this paste is awesome. Temps are very good. It's nice and think consistency. Easy to work with. All around perfect performance. Also have arctic 5 used it for years also good lol

Read some reviews with the Noctua paste. Always high in the tests.

 

[freeagentt]

I think AS5 is still ok for lower end builds. Its pretty stable once it sets and it lasts forever. I find once you start scootin past 180-200w it loses its grip a little.

I haven't tried any MX series pastes yet..

 

[DWD1961]

I think AS5 is still ok for lower end builds. Its pretty stable once it sets and it lasts forever. I find once you start scootin past 180-200w it loses its grip a little.

I haven't tried any MX series pastes yet..

I think it's great for high end builds too, but it gets shorted in the 'reviews.' It does last forever and it's shelf life is virtually forever. But I'm curious why you would think wattage has anything to do with it. Isn't a paste's function temperature? A 200 watt CPU can run cool, right, and they all have basically the same max temps you need to keep it below, like around 80C? Did you mean higher heat generating applications is where it breaks down? It never let me down OCing the first gen Core i7 920 at 4Ghz -- on air. If I remember correctly, that 4Ghz OC was pushing 200 watts.

 

[Starfalcon]

Now I feel really old, if arctic silver 5 is considered ancient....I still have some tubes of arctic silver 2 and 3...

 

[freeagentt]

I think it's great for high end builds too, but it gets shorted in the 'reviews.' It does last forever and it's shelf life is virtually forever.

Yeah its pretty decent. Not the best though, trust me I swore by the stuff for at least a decade.

But I'm curious why you would think wattage has anything to do with it. Isn't a paste's function temperature?

I should have been a little more clear, passing 200w, and approaching 250w on 7nm AMD is quite intense.. And under 180PPT it was fine on my 5900X, but it lost its grip at around 200w PPT. I have pushed old Intel CPUs way past 200w, back in the day 200w was just getting started, and I used AS5 back then. I am pretty sure my X5690 was able to do 350w.. according to my meter.

 

[freeagentt]

Hey, I just tried some Ceramique2 for the first time in god knows how long, and I liked it lol. It was not bad.. Not the best, but heck it was good to see Arctic Silver still in the game

 

[xDiVolatilX]

Read some reviews with the Noctua paste. Always high in the tests.

What do you mean? lol

 

[DWD1961]

Hey, I just tried some Ceramique2 for the first time in god knows how long, and I liked it lol. It was not bad.. Not the best, but heck it was good to see Arctic Silver still in the game

I'm going to test Artic Silver 5 though, not their ceramic stuff. I think I actually used that once, way back when horses were pulling carts, or so it seems.

 

[DWD1961]

What do you mean? lol

Reviews rated Noctua paste very high. It's very good. As reported below, the Arctic Silver 5 still performed well at only 4.6C higher than the best tested paste (ProlimaTech PK-3), not including the liquid metals. Next best nonLiquid metal was the Noctua at 42.88, which is only 3C better than Arctic Silver 5. (If you believe their test results.)
https://www.tomshardware.com/best-picks/best-thermal-paste

 

[DWD1961]

Results are up in original post above.

 

[pendragon1]

thanks!
how did the paste spread look between changes? it looks like it might have been too little paste.
results seem like what ive always found; if youre not looking for every last degree c, as5 is still fine.

 

[E4g1e]

Now I feel really old, if arctic silver 5 is considered ancient....I still have some tubes of arctic silver 2 and 3...

AS5 is considered "ancient" due to its extremely long required curing time, not because of its performance per se. It requires a minimum total of 200 hours (more than eight full days) of curing (this means intermittent on and off cycles, with no heavy loads whatsoever on the CPU) before it will perform at its best. Other thermal pastes are ready for full performance in as little as 30 minutes.

 

[xDiVolatilX]

Reviews rated Noctua paste very high. It's very good. As reported below, the Arctic Silver 5 still performed well at only 4.6C higher than the best tested paste (ProlimaTech PK-3), not including the liquid metals. Next best nonLiquid metal was the Noctua at 42.88, which is only 3C better than Arctic Silver 5. (If you believe their test results.)
https://www.tomshardware.com/best-picks/best-thermal-paste
View attachment 536385

Wow no wonder it gets such high remarks. And the large tube was very reasonably priced. Glad I got one. It may be my go to forever from now on. Previously always used arctic silver 5 for years and years. But Noctua NT-H1 is the new kid in town lol it's nice thick yet buttery smooth perfect consistency imo.

 

[DWD1961]

thanks!
how did the paste spread look between changes? it looks like it might have been too little paste.
results seem like what ive always found; if youre not looking for every last degree c, as5 is still fine.

It looked like a circle spread all the way out to the edges. Just a little paste was on the sides of the CPU. I was going to take a picture, but decided not. Most people put WAY too much paste on so it squishes out over the sides the the CPU. I have paste application down to science. A couple of years ago, I also tried every application method stated and found none better than just a rice size grain right in the middle (a small rice grain).

However, from my application and testing expereince, I decided to go with the 5 point procedure to allow the paste to spread out better without putting so much right in teh middle. It's just easier for me to judge how much I need to cover, but not exceed, the heat spreader--by much anyway.

 

[pendragon1]

It looked like a circle spread all the way out to the edges. Just a little paste was on the sides of the CPU. I was going to take a picture, but decided not. Most people put WAY too much paste on so it squishes out over the sides the the CPU. I have paste application down to science. A couple of years ago, I also tried every application method stated and found none better than just a rice size grain right in the middle (a small rice grain).

However, from my application and testing expereince, I decided to go with the 5 point procedure to allow the paste to spread out better without putting so much right in teh middle. It's just easier for me to judge how much I need to cover, but not exceed, the heat spreader--by much anyway.

the pattern is good, just need slightly bigger dabs. theres no such thing as too much, other than making a mess.

 

[DWD1961]

AS5 is considered "ancient" due to its extremely long required curing time, not because of its performance per se. It requires a minimum total of 200 hours (more than eight full days) of curing (this means intermittent on and off cycles, with no heavy loads whatsoever on the CPU) before it will perform at its best. Other thermal pastes are ready for full performance in as little as 30 minutes.

I've seen a test where they tested AC5 right after putting it on and then actually giving it cure time. It was a bout a 1-2C difference. I allowed the AC5 I used to cure, but I also got it hot a few times. I have not read, or have forgotten that they said to NOT heat it up at all during cure time. It just says let it cure.

"Due to the unique shape and sizes of the particles in Arctic Silver 5's conductive matrix, it will take a up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink or with a low speed fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop 2C to 5C over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired."
Arctic Silver 5 Website

 

[DWD1961]

Wow no wonder it gets such high remarks. And the large tube was very reasonably priced. Glad I got one. It may be my go to forever from now on. Previously always used arctic silver 5 for years and years. But Noctua NT-H1 is the new kid in town lol it's nice thick yet buttery smooth perfect consistency imo.

Next time I might use that too, or I may get a tube and post the results here also?

 

[DWD1961]

the pattern is good, just need slightly bigger dabs. theres no such thing as too much, other than making a mess.

When I took that application off, it was the AC5 application, it went all the way to the edges with just a little over run around the edge of the heat spreader. I'm calling that a pretty near perfect application. However, I did apply just a little less for that specific application. Still, it ran out on the edges.

This was from my last heat test back in 2020 when I thought my 3600 was running hot. The newest application looked exactly like this one.

 

[pendragon1]

you need a tiny bit more to cover the corners. unless you heatsink doesnt cover them, then you need a new heatsink... could also use a bit more pressure, looks kinda thick in the middle.

 

[DWD1961]

you need a tiny bit more to cover the corners. unless you heatsink doesnt cover them, then you need a new heatsink... could also use a bit more pressure, looks kinda thick in the middle.

The other 'part' of the paste was on the block. It never comes off equally on either the block or spreader. As stated above, I tried also spreading it with a credit card corner to corner and no difference in temps.

I literally tried every spread application I read about and the center rice sized grain proved, in my application, no better or worse than any of the others -- exact same temps.

 

[pendragon1]

The other 'part' of the paste was on the block. It never comes off equally on either the block or spreader. As stated above, I tried also spreading it with a credit card corner to corner and no difference in temps.

I literally tried every spread application I read about and the center rice sized grain proved, in my application, no better or worse than any of the others -- exact same temps.

again, not arguing the pattern, just the amount. you should have paste corner to corner and making the outside dots just a tiny bit bigger is all thats needed.

 

[DWD1961]

again, not arguing the pattern, just the amount. you should have paste corner to corner and making the outside dots just a tiny bit bigger is all thats needed.

Did you watch the video? He gets the same spread in the corners as I did vs a corner to corner full spread, and he is using an AMD 3700X -- and no difference.

I dunno man. I can easily do it again but it won't matter.

 

[xDiVolatilX]

Next time I might use that too, or I may get a tube and post the results here also?

I didn't take note of any temps because my information and testing isn't extensive or proper enough for me to consider posting it here as public information but I will say that it isn't very far off from the conductonought liquid metal I used prior to it. I didn't notice much if any a rise in temps. The difference is negligible. The Noctua NTH1 paste is so good that I will never use liquid metal ever again not that it's even meant to be used on an IHS anyway. I have never owned a Noctua air cooler so 0 fanboyism because I've been on water for a long number of years since the ol days when Zalman was the king of air coolers but I will say this Noctua paste is VERY impressive. All those thousands of 5 star reviews are %100 true and I am now a firm believer in the Noctua paste. It's likely the last paste I'll ever use forever It probably doesn't or can't get much better than this.

 

[xDiVolatilX]

again, not arguing the pattern, just the amount. you should have paste corner to corner and making the outside dots just a tiny bit bigger is all thats needed.

I could never do the pea or dots or X pattern method When I applied the NTH1 paste it was fully covered %100 I have always and will always do it this way I don't care if I waste any because I repaste very few and far between and paste is so cheap I don't mind being as generous as I need to be. I don't leave any metal of the IHS exposed lol it is a thin spread like peanut butter like fully covered so satisfying haha

 

[THRESHIN]

what i miss is the arctic silver alumina epoxy. two part ceramic thermal epoxy. nothing gets that shit off.....EXCEPT!!!!! mix up a wad of that stuff and then mix in an equal part of AS5 paste. you'd get a cement for attaching heatsinks to stuff like GPU memory that was very removable. almost rubber like coming off.

nobody sells that shit anymore. i'm going to go cry myself to sleep now.

 

[Starfalcon]

what i miss is the arctic silver alumina epoxy. two part ceramic thermal epoxy. nothing gets that shit off.....EXCEPT!!!!! mix up a wad of that stuff and then mix in an equal part of AS5 paste. you'd get a cement for attaching heatsinks to stuff like GPU memory that was very removable. almost rubber like coming off.

nobody sells that shit anymore. i'm going to go cry myself to sleep now.

I have some of that laying around too, in the bag with my arctic silver 2 and 3.

 

[THRESHIN]

I have some of that laying around too, in the bag with my arctic silver 2 and 3.

It's probably turned to stone by now. That was always the problem I had with it. Still, cool stuff. I don't keep much of my old crap around.

 

[philb2]

It's probably turned to stone by now. That was always the problem I had with it. Still, cool stuff. I don't keep much of my old crap around.

The really old stuff, I just toss. And I always buy smallest possible quantities of paste, because I'm not a guy who is constantly changing out the cooler, etc. My ideal paste is set-it-and-forget-it until I upgrade with a new cpu and maybe motherboard.

 

[freeagentt]

Give it a little squirt on some paper or something and if water comes out first then it has separated. I have seen year old AS5 that only had one squirt removed do that.

 

[Starfalcon]

It's probably turned to stone by now. That was always the problem I had with it. Still, cool stuff. I don't keep much of my old crap around.

Yeah it likely has by now I am sure, got back in the day to heatsink the memory on my 9500 pro. I ended up getting a 9800 pro shotly after that so I never did use it.

 

[freeagentt]

Yeah that's only a few years old 😜

 

[kirbyrj]

If you don't have your graphs centered at zero, it's a useless graph. The baseline isn't even consistent between idle, load, etc.

See #3 on this website...it's an old tool used in Lying with Statistics.