Arctic MX-5 vs MX-4 vs MX-2... What's the difference?... Also Noctua NT-H1 

Had a blast watching this, great sense of humor :D

Slow and steady, that what I keep telling myself anyway :D It would be great if this channel took off, I think I just need to manage it better.

yes and why?? this video are crap

Yeah I think there's some scope to do more of this in the future. I'll probably drop the "spread" testing since I don't think it's that important, so with that being the case I could probably get a couple of these video completed for every other kind of review I do. Should be good to balance the upload schedule.

i remember mx2 and mx4 had both one is more liquid the other is more thick (and annoying to remove)

I had the opposite results, my MX4 changed drastically over the course of time, even when standar intel/amd paste or even regular silicon paste was much more stable

I’ve heard about people having pump out issues within a week with the Noctua ones and ended up throwing out the Noctua. Just ended up ordering some MX-4 as it’s the tried and true and it’s what I used years ago. Don’t fix what isn’t broken! I’d rather have years of longevity rather than a week.

But how often are you trying to run your CPU at temps higher than 65c?

The CPU would probably hit 100C if the paste wasn't there, so it's the paste forming a good transfer between the two that's allowing it to happen. I could test with a smaller cooler which would it increase the temperature, but frankly I think maxing out at 65C is a pretty reasonable testing environment for the paste to adequately prove itself. On the flip side, a paste that performs alright at 65C, but really well at 90C is a bit of a poor strategy. Surely you'd want one to perform well enough to encourage the temperature to not hit 90C in the first place (all else being equal). This is all assuming the properties of the paste change significantly between 65C and 90C, which I don't think is the case as you'd probably class it as an unstable material. I'm only saying that since I don't know enough about material science to know of a material that significantly increases thermal conductivity, unless it's changing states. Hopefully that kind of covers my testing methodology, but I can see why maybe testing with both a small and a big cooler to simulate less and more challenging environments would be useful. It would also be time consuming to do which is why I've landed on this method being good enough. But if I do this in the future, and there's some indication that there's a significant performance advantage at a higher load, I can have a play around and see what comes of it. Regarding the themral paste application. I've been doing this method for all the CPU cooler testing, and presenting how it spreads, and it's good enough for the vast majority of those. The main reason it shouldn't spread out well is poor mounting pressure. But since we're using the same cooler for all three, it's on a relatively level playing field.

@@headerahelix Well if your cpu runs at just 65c, heat isn't an issue so any thermal paste would do the job. When thermal performance of a paste becomes relevant is when heat output is relevant. I'd say around 85c maybe? Then heat becomes relevant in terms of how high clocks you'll reach.

You do understand that the thermal paste is part of the cooling solution that enables the CPU to achieve 65C in the testing right? If it performed worse, then it would get hotter, of it performed better it would be cooler. And to be fair, it's 65C delta T, so it's true temperature is about 20-25C (room temperature) above that. Around 85-90C, about as hot as I could get it before thermal throttling. I could have used the channel standard Mugen 5, but I wanted to push the paste and see what it did at the extremes. Maybe it would have been good to also test with the Mugen 5 and drop down to 65C true temperature as well.

Honestly, any will do. I wouldn't stress over it. I've used MX-2 for my GPUs which works well, but any other will work.

@@AVTechy I bought it after you said it’s fine. It is - if you’re cool with changing it every 2 to 4 months.

Silver 5 is awful and should not be considered.

@@CknSalad strange you would say that , I believe this product has by far the most positive reviews out there.

What results did you get with Mx4?

@@Interestingworld4567 you sure you’re responding to the right comment bro? I’ve used the same vial of MX4 since 2016, works fine, haven’t compared in a controlled way to any other TIM.

Thanks for highlighting this. Actually first I'd heard of the issues, good to be aware of.

@@AVTechy You're welcome. I have a new PC with 5800x and dark rock Pro 4 and 4x8gb 3600mhz cl16 rams. I built it myself. It didn't boot and the dram led was constant red. That could be a lot of different issues. More likely ram or ram slot problem. I couldn't take my cpu cooler off. I need to take off the dark rock Pro 4 because I wouldn't be able to check the rams otherwise because of it's size. so I took it to the shop where I bought all my parts except the phanteks p500a case and my gpu. They had to rip off the cooler and the cpu came with it. Literally, it was underneath the cooler like glued to it. Fortunately, no bent or broken pins. It turned out, my problem was this. I bought a b550 tomahawk motherboard which had the label on it, it is ryzen 5000 desktop ready. Well, it wasn't. All I needed was a bios update. Which I wanted to do first, by just plugging the psu with the bios usb. But I didn't want to bother since it should've been worked. Now my pc is working and they were able to put it back and repaste it, unfortunately with the same paste. I told them I would take it to them to repaste it in the future. Dangerous game this paste is.

Wow... Was about to order some as it's on clearance lol I think actually, this would be an excellent thermal glue then for gluing heatsinks into VRMs and memory modules. I guess I'll still order but until I know it's safe for CPUs just use it for GPU heatsinks.

Not sure it's even wroth it for GPUs since I don't think we know what the performance is like once it's separated. It could have uneven thermal properties across the die, and lead to performance issues as a result (pure speculation). I'd recommend going for something that hasn't got a manufacturing issue, just to be safe.

@@headerahelix I agree whlith AV techy in this matter. I would go for mx4. It's still a very good paste and cheap. If money is not an issue, go for grizzly kryonaut. They are both trustworthy and good pastes.

Thanks very much man, really appreciate that :)

Take those numbers with a pinch of salt. If you're really pushing your system regularly you'll probably want to factor in a thermal paste change after a year or two depending on a lot of factors. Otherwise you can get more life out them. But it all depends on what you're using it with, how hot it gets, how regularly it hits high temperatures and for how long, etc. In reality most decent pastes will last at least a few years, but after that it's dependant on how you're using it.

Small addition, I've used MX2 with my test GPU for years now, and it's still performing well. Don't get too bogged down with the details. Get something decent and don't worry about it too much.

And every tech person on here recommends to NOT spread it and just apply a small pea sized amount. Go ahead and throw that spreader away lol

@@allieboy181 I don't care what they do. I want to know that I have full coverage. I've seen too many cooler plates pulled off with sections of the cpu that never had any thermal compound because of the blob and squish method. I'll take the time to do it right and not be a lazy ass with my system. You do what you want with yours.

@@RedPill780 Because it's easier to do a thorough cleaning of my air cooler/fans if I remove it which is usually easy because the thermal compound hasn't turned into epoxy. You do your maintenance the way you want and I'll handle my system the way I wish.

just marketing 8 year later who use it again without buy new cheap one

they say you get the best results over time but well, try putting back again mx2 and see if you get the same 75c tops.

Any updates?

MX-2 seems to have bad endurance from what I can tell

the MX-4 seems to be the best option...

I have noticed many "thinner" very flowy, "thats a technical term", thermal pastes like MX-5, kryonaught do very well with very flat, smooth surfaces, and high mounting pressure, but do worse in less then ideal surfaces. With rougher surfaces or non flat surfaces i have seen better results with thicker pastes like AS5 , mx-4.

noctua nt-h2 slightly better, but these are two excellent thermal pastes with strict production quality control

Looked kinda shifty to me. Can't trust a guy with rolled up cuffs...

Sounds about right.

Yes, that is the topic of the video.

@@AVTechy but where’s the answer

one is very liquid and easy to clean, and the other is pretty thick and makes a mess to remove cause its thick.. that is all i got from it having both.. mx4 and mx2

I have MX-4. I believe it is not as thick of a compound. While it may have got the short end of the stick on this review, I'm wondering if the compound being thinner will help it last longer than the other n prove why it is, in actuality, an upgrade.

Looks like I got 'dissed'!

No idea, he just climbed through the window one day and got really into testing stuff. I'd kick him out, but then I'd need to do the testing again...

Not my experience, but fair enough.

@@AVTechy Happened a few times with me with CPUs and GPUs that ran hot. After a year or 2 the temps got higher and higher and when I reapplied new paste the old one was so dry that it flaked off in clumps

I'll take it :D

yeah I don't understand how you'd spend so much time putting together a nice video only to not normalize the audio...

Nice!

I'm not really saying anything, the results look like that's the case. There's not a lot in it either way to be honest.

@@AVTechy that's weird, usually mx-4 beats mx-2.

He's saying it doesn't really matter, you're best off buying the most cost effective paste you can for the performance you get.

Last time I did a head to head between MX-2 and MX-4 they were equal. This testing was a little more in depth, and it looked like MX-2 pulled ahead slightly. We're only taking max 2 degrees here, so it's not substantial. Like I said in the conclusion, if one of them goes in sale for a couple of dollars less it will pull ahead in value. So yeah, it really comes down to getting the cheapest one.

@@AVTechy Few thermally conductive fasting gives positive results immediately after application and sure tests will be repeated one month after application. Arctic thermally conductive pastes have the phenomenal property that they have greater thermal conductivity over time.

I'm still yet to work with that stuff. Like Noctua, it's got a reputation for being really good. Maybe next on the list for thermal paste testing.

It's also the most expensive, but similar in performance and price to noctua nh-2. Arctic MX-5 is said to be on par with these two performance wise but I think it's slightly cheaper.

@@AVTechy And Kingpin Cooling KPx please! It has great reputation in where I'm from, along with Kryonaut, NT-H1, and MX-4.

maybe a little bit better while costing more and being conductive so u have to be more carefull while applying it.

@@nicane-9966 Kyronaut is not conductive, neither is kryonaut extreme, hydronaut, or aeronaut. Only their liquid metal conductonaut is electrically conductive, obviously.

Hence the comment at the end of that section.

Are you talking about 2 or 4? I've had 2 on the test system GPU for years and it's still performing really well. I haven't endurance tested 4 yet, so I'm not sure about that one.

@@AVTechy I'm talking about 2, had no problems with 4 but lost it when I moved

Ah okay, just noticed you said you were disappointed with 4 at the end, just wanted to clarify.

@@AVTechy edited

you bought a fake one bro

No, your certainly not. The differences are minimal, but testing confirms whether there are or are not differences. Typically anything mid range will perform similarly, but for all I know a new to market mid range paste could perform good enough to make it a standout winner. We'll only know this with testing.

@@AVTechy difference is mostly like measuring error, you even dont really know what is better, its more abou put it correctly on cpu than about what exact paste it is

Measuring error could be up to +/- 1 degree. Outside of that were likely looking at difference in performance. Other reviewers have lower margins of error than that, but I don't think I'm on that level despite doing my best to be as consistent as possible. You're right, the differences are minimal, but if we test and confirm this then we can conclude that the cheapest on will do. What I'm getting at is assuming over lots of years of using paste that they're all the same so it doesn't matter is at best an assumption, so confirming that, or otherwise, is important. Not trying to stir the pot, but I have a feeling you weren't 100% on this, otherwise I'm not sure why you'd have engaged with this video.

Good job we normalised for all of that I guess.

Yeah, heard about that. Turns out they've had a really bad batch, and I'm not sure how QA didn't pick up on it considering how quickly everyone else did. When it's made properly it's alright, but this has certainly hit its reputation.

thank for sharing all that, I was already thinking of getting mx4, your experience helps affirm mx4 is likely the best and ideal choice