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

[ir_cow]

It's not. For starters, it's grey not blue. I'll detail more in my review this weekend.

Ah yes, removing the blue dye will fool them all ahha. I'm just saying mx5 was recalled / discontinued because of bad press performance issues in some batches. There is no coming back from that. They did exactly what I would do. Remove it from sale for a few months and rename it.

 

[lexluthermiester]

Ah yes, removing the blue dye will fool them all ahha.

Well I haven't tested it yet, but that's a start.

I'm just saying mx5 was recalled / discontinued because of bad press performance issues in some batches.

I never had problems and bought a bunch of it. People over-react ALOT!..

There is no coming back from that.

I disagree. Arctic is a company that makes quality products. Just because they had one problem doesn't mean the whole company is crap. They made good on returns, exchanges and swap outs. To me, that speaks well of them.

It think they should have stood their ground, and told everyone: " Hey, we had a bad batch, it's fixed. Let's get on with things." and then continued to make it.

They did exactly what I would do. Remove it from sale for a few months and rename it.

The testing will tell the story. If the consistency is similar and the performance matches, then that might have been what they did.

I'm curious myself. If they did just rework the formulation and change/remove the colour, it would still be a good value as the performance of MX-5 was excellent!

 

[Space Lynx]

MX-6 was just good timing for me, not only can I use some now, but I know its fresh out of the factory and will legit last me 8+ years. Well, most likely it will. I plan to build this PC and just leave it with moderate dusting every 3 months. Building computers has become too stressful, I'm ready to be done. Now I just need to score me an RDNA3!

 

[ir_cow]

I still have 2 unopened mx5 tubes that was sent to me. The first one was okay and seemed to be within 1c of the thermal grizzly. I don't like the gummy part. Hard to clean up and I'm repasting upwards of 5-6 times in one day sometimes.

 

[lexluthermiester]

I still have 2 unopened mx5 tubes that was sent to me. The first one was okay and seemed to be within 1c of the thermal grizzly. I don't like the gummy part. Hard to clean up and I'm repasting upwards of 5-6 times in one day sometimes.

Fair enough. I imagine frequent repasting would get annoying.

 

[Calenhad]

I have tried a graphite pad a couple years ago, and I got 3-5 celsius worse than paste. probably doesn't matter to most people, but I want my CPU to stay as cold as I can get it. hence this thread really, just fun hobby to try and find the best paste. at end of day though, it doesn't matter.

Not saying graphite pads are the new best thing. But for longevity they are an interesting option. No idea how those new Thermal Grizzly pads are compared to older ones. But if what you want is a no maintenance cooling solution, that will last the lifetime of the computer, then it could be worth a couple degrees.

My personal computers all have custom watercooling and I use KPx paste. So I am definetly not the target customer for graphite pads either. But I build enough computers for non-enthusiasts, such that a no maintenance solution is intriguing

 

[MachineLearning]

I still have 2 unopened mx5 tubes that was sent to me. The first one was okay and seemed to be within 1c of the thermal grizzly. I don't like the gummy part. Hard to clean up and I'm repasting upwards of 5-6 times in one day sometimes.

MX-2 is probably the easiest to clean up in my experience.

Don't ever try Arctic Silver Ceramiqué 2 for benching. Thickest paste I've ever used.

 

[jayjr1105]

I'm going to hold off for now. I still have a few tubes of GD900, plus a small amount of MX5 and Master Gel left. I'm hoping the MX6 can bridge the gap and get close to Kryonaut.

 

[lexluthermiester]

Ok, the OC phase of testing for Arctic Cooling's new MX-6 is done. I ended up testing it against Arctic's own(and now discontinued) MX-5, Noctua's NT-H1 and CoolerMaster's CryoFuze Violet. The results were interesting.

Had to change the hardware config of the tests to get a stable OC from the W3680 in the Dell T3500. Hyperthreading, Speedstep and C-States were disabled in the BIOS. ThrottleStop 9.00 was used as newer versions do not work as well with my T3500. Not sure why. But Thanks to @unclewebb all the same for making such an amazing utility! In Throttlestop, the TDP was increased to 150W from 130W and the TDC(A) was increased to 130A from 115A. The multiplier was set to 31x133mhz for a final OC of 4133mhz, the best I can get on the system. A full system shut down and cool-off was done after every test to ensure no residual heat from a previous test was retained.

Here we go. First up are the Prime95 runs.

Since I had Cooler Master CryoFuze Violet already installed from the last series of testing, it came first.

Next was Arctic MX-6.

Next was Arctic MX-5.

Finally we have Noctua NT-H1.

These tests were done with a minimum of 5 minute runs, but I was curious about the temps and how low they were so a few tests were longer than others, which didn't seem to have any effect. However, the results are valid within the context of these tests. These numbers are still a good way to compare the the TIM's in question to each other.

Next up we have the CineBench R15 Extreme runs.

Again we have CMCFV.

Next MX6.

Next MX5.

And finally NTH1.

It seems Cinebench just doesn't stress the CPU as much as Prime95, which is why the temps didn't go as high. This was expected. Cinebench R23 did not want to run, perhaps because of Windows 7 or lack of AVX on the Xeon.

So as you can see, Arctic MX-6 is a very worthy followup and improvement to MX-5, which despite the batch problems and bad press, was an excellent TIM. I'm going to do another round of tests with all normal settings(no TS OCing) over in the Official TIM thread in a few days.

My personal take? I was very excited about MX-5 last year and MX-6 is even better. Nominal config system testing still needs to be done, but so far so good.

 

[Space Lynx]

I have got my rig built. Just waiting on gpu now. I did use MX-6 though, so its good to hear this! Thanks for the results @lexluthermiester

I am using the LGA 1700 Contact Frame, genuine version from Thermalright. So that should help me a little bit too.

Looks like one of my 200mm fans is dead on arrival. This case has been sitting in its box new for a year or more now, I never used it until now. SO that is a bummer, but luckily, I have two spare 120mm that will replace it... and allow me to keep the other 200mm on the bottom, so eh its all good. Main thing is keep that air moving.

 

[maxfly]

Great job Lex!
I'm getting there, ran into some roadblocks but hoping to get at it tomorrow afternoon or evening.

 

[Divide Overflow]

I bought 4g of MX-6 with the cleaner pads online from Amazon. After delivery today I'm somewhat skeptical of Arctic's authenticity check. The printing on my packaging was badly misaligned. I've contacted Arctic's customer service to report this defect. Is anyone else seeing something like this?

 

[erocker]

I'd return that. The QR code being messed up isn't good either.

 

[freeagent]

The grey coating isn't very thick, is it? You almost don't need to rub it off.

 

[jayjr1105]

I bought 4g of MX-6 with the cleaner pads online from Amazon. After delivery today I'm somewhat skeptical of Arctic's authenticity check. The printing on my packaging was badly misaligned. I've contacted Arctic's customer service to report this defect. Is anyone else seeing something like this?

View attachment 269981

There is no way someone has already made a counterfeit copy lol. It's genuine dude, just scratch off the gray and scan

 

[freeagent]

What has me concerned is the temp spread between his cores. Thinking back to my x5690 running an H100 and AS5 I would see a 3-6c spread while using a higher vcore and being clocked a few hundred MHz higher. But I believe his spread is due to the heat sink itself.. but I could be wrong.

 

[lexluthermiester]

Is anyone else seeing something like this?

Looks ok to me.

I'd return that. The QR code being messed up isn't good either.

It's a scratch-off QR code.

It's genuine dude, just scratch off the gray and scan

This.

What has me concerned is the temp spread between his cores. Thinking back to my x5690 running an H100 and AS5 I would see a 3-6c spread while using a higher vcore and being clocked a few hundred MHz higher. But I believe his spread is due to the heat sink itself.. but I could be wrong.

It's the IHS. The W3680 is not perfectly flat and I don't want to lap it. The heatsink is perfectly flat, but has a rough surface. Perfect testing case scenario for TIM's, which is why I use it. If a TIM performs well on THIS system, it will perform well on any system.

Below is the spread after removal;

CMCFV after testing.

MX6 after testing.
As you can see, the heat sink surface is rough but perfectly flat and the imprint of the IHS is uneven, but fortunately center-line.

 

[Divide Overflow]

Looks ok to me.

Not my question, but thanks for your opinion.

It's a scratch-off QR code.

It's an authentication process that is badly compromised by poor packaging in my sample. I was curious if anyone else was seeing something similar.

I've been pretty happy with my early batch of MX-5, although it does appear to be very gummy after use for a few months. I'm sure I'll be equally happy with the MX-6 revision.

 

[maxfly]

Not my question, but thanks for your opinion.

It's an authentication process that is badly compromised by poor packaging in my sample. I was curious if anyone else was seeing something similar.

I've been pretty happy with my early batch of MX-5, although it does appear to be very gummy after use for a few months. I'm sure I'll be equally happy with the MX-6 revision.

If the scratch off is intact your gtg. My box is semi squashed on one end but the tape held it together. The QR scratch off on mine looks the same as yours. Not sure what your concerns are?

Alright, started off with Noctua Nt-H1 and swapped in the U12A for simplicity. The PA120se is bigger and harder the move around. Was a little slow getting HWINFO started with P95...Ima bit rusty. Idles were 26-27c across the board sorry. Load temps are where the meat n taters are at anyhow hehe.

And on to MX-5 next.

And finally MX-6

Ambient temps were kept at a steady 20c throughout testing.

Ive never used NT-H1 before this run and was I really surprised at how it glued itself to the friggin cpu! I had to reef on the hs to get it off. I couldn't twist or turn it for anything, I had to pull it straight up until it finally popped off. This is while the cpu was fresh off a p95 run so it was roasty, toasty at that, so something to keep in mind AM4 owners. Temps were fine of course but dayum that stuff didn't want to let go!

The tube of MX-5 I have is no different than the first time I used it initially. Very soft and the easiest to spread by far. Removing the hs of course leaves little melted cheesie like strings that can be a pita to chase around if you arent anticipating them...or have forgotten that it happens with this TIM. Temps were fine once again.

MX-6 was the thickest of the three and hardest to apply in a thin layer. It actually dropped the idle AND load temps which is unusual so I ran p95 a couple of more times for 9-10 minutes each just to see if there was something wonky going on. Nope nothing changed. So yeah, its solid stuff. Ill keep it on until I get a wild hair and or energetic enough to either swap heatsinks again or put PK3 back on. If I had to choose between these three and PK3 I would have a hard time choosing between MX-6 and PK3, I would probably still go with PK3 only because I like how easy it is to apply vs how much thicker the MX-6 is.

 

[Space Lynx]

Well, this is great news. Really glad I went with MX-6 on this new build of mine. I like that its fresh from the factory too, I know I can leave it for 8 years and not worry. Hell, my dads all in one pc is ten years going strong and still runs fine on temps.

I found putting on the MX-6 to be more difficult than I was expecting too because of its thickness, so nice to know I am not alone in that regard, but it wasn't that big of a deal.

MX-6 + thermaltake contact frame + the extra fan in push/pull configuration instead of just leaving the one fan on the air cooler... its prob only 2-3 celsius gains for each of those alone... but together i am looking at 8-11 celsius better temps overall...

it adds up... that's for sure. of course this situation only applies to LGA 1700 socket with the contact frame, but yeah, that extra variable x3 adds up quite nicely. I will be surprised if my raptor lake breaks 70 celsius in witcher 3 max settings 1440p 165hz 165 fps... we will see... that is going to be first test I do, let it run for like 2-3 hours. just to test system stability. I will ride my horse into the big cities as well for most of that time, really load up the cpu that way.

so glad I didn't go with AM5 for this 5-7 year build... I have a weird thing when it comes to high temps, it just bugs me to no end. can't stand it really. so seeing 95 celsius often would have made me insane. lol

 

[lexluthermiester]

Not my question, but thanks for your opinion.

Might have misunderstood me there. I meant that the QR code looks like it's legit.

Curious about viscosity.

MX-6 was the thickest of the three and hardest to apply in a thin layer.

Should probably have touched on that. MX-6 is a thick, slightly dry paste. Took some effort to spread! Maxfly and I seem to have had a similar experience on that point.

I know I can leave it for 8 years and not worry.

Yeah, that is likely going to be ok.

 

[Space Lynx]

If you want we can move the convo to the official thermal paste thread. I think we have our answers? I'm ok if mods want to lock this thread, I have no idea why a thread on mx-5 and mx-6 reached 170k views, but here we are. I will leave it up to you. I am ready to game though, less testing, more gaming. Nice to know I got my paste sorted now though, lol

ty for all the testing everyone contributed to this thread.



(if you all still have the thread open when my rig is done, I will post my witcher 3 results with this paste) I usually pick the same spots in cities for my testing, so its still consistent results.

 

[MachineLearning]

Should probably have touched on that. MX-6 is a thick, slightly dry paste. Took some effort to spread! Maxfly and I seem to have had a similar experience on that point.

Good to hear. Probably bodes well for longevity, which is what I'm most interested in at this point (since it can't yet be answered ). I prefer thicker pastes either way.

 

[lexluthermiester]

If you want we can move the convo to the official thermal paste thread. I think we have our answers?

Agreed. As stated above, I'm going to post the nominal clock based tests over there.

I have no idea why a thread on mx-5 and mx-6 reached 170k views

I think it was a combination of interest, excitement and then the controversy.

ty for all the testing everyone contributed to this thread.

You're welcome! It was fun!

Good to hear. Probably bodes well for longevity, which is what I'm most interested in at this point (since it can't yet be answered ). I prefer thicker pastes either way.

Agreed. It was challenge to spread but not really super difficult. I believe MX-6 might become the next AS5 in terms of longevity and performance/value. I had that hope for MX-5, but...

 

[MachineLearning]

Agreed. It was challenge to spread but not really super difficult. I believe MX-6 might become the next AS5 in terms of longevity and performance/value. I had that hope for MX-5, but...

I seem to recall you working at a PC repair shop, do you often deal with laptops? Would you ever have a chance to test MX-6 in such a situation?

Currently use MX-2 in my laptop because it performs better than MX-4 does in that use case, and I would replace it if something else is significantly better.