Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Best Thermal Paste?

mac57mac57

mac57mac57

macrumors 6502a
Original poster
Aug 2, 2024
524
312
Myrtle Beach, SC
I realize that this may cause more than a few responses since it is an opinion question, but as I embark on air cooling a G5 Quad, I would like to solicit the opinions of those in the know here at MacRumors...

What is the best thermal paste to use with a Quad? By "best" I would guess that the criteria likely centers around it's thermal conductivity.

Thanks in advance!
 
I used Arctic MX-6 in the PC Cooler Quad.

You probably want to look for TIM that can do at least 10 W/mK.
 
In almost all of the hardware forums I've come across (which admittedly is not too too many as I'm more of a software guy), the almost unanimous opinion seems to be Arcitic's thermal paste.

However, there are also various opinions about what is the best pattern. There is a really good video series of a guy who does a lot of various testing with thermal paste, which you may find interesting:

www.youtube.com

mryeester

I tinker with tech! Business email: yeester12@gmail.com www.mryeester.com www.yeesterpaste.com Thanks for stopping by! Owned and operated by Adam Yee @mryeester
www.youtube.com www.youtube.com
 
Excellent, thanks all! I will review the videos and see what insights they provide. In the meantime, my web research is pointing me towards Thermal Grizzly Kryonaut Extreme as "the best of the best". Thoughts?
 
I stick with MX-4 and have seen no reason to change. Pastes are much of a muchness and you are unlikely to shave off more than an degree or so, whatever you pick. I tried Grizzly Kryonaut and it was frustratingly unspreadable. It also does not have great endurance so you will have to repaste every other year or so. If you have a large die, look elsewhere.
 
I'm a creature of habit and generally use Antec Formula 7 ND. It is what I use in most of my macs. I have a 4g tube of ArcticSilver 5 too but it is thinner and easy to make a mess with, so kind of a pita to use. I did use it on my sons imac when I repasted its cpu/gpu and Im getting good thernals with it - not so much because I prefer it but rather because my AF7 was out in the garage lol and AS5 was what I had around in the house but it is doing its job.

I have not used Grizzly kryonaut before so dont have a personal opinion of it.
 
Artic Mx4 is what I've pretty much always used. It seems like it flows well and I feel I can get good coverage with it which I feel is more important than one or two C here and there. I'm pretty sure it's non conductive too which is another reason I don't bother with the fancier stuff. I re pasted my dual core 2.3 GHZ G5 and all it's other thermal paste needing doodads with it and it seems to be alright. But who knows, I haven't checked the temps or used it all that much since I did. What a painful case design. I think it would have been far better if the case had two removable panels and if the board was attached to the rear one so it could just pop out. It was a nightmare for me at least getting everything out from that one side. I wish you luck on your quest!
 
  • Like
Reactions: Certificate of Excellence
Artic MX-4 has a thermal conductivity of 8.5 W/mK. This is a good rating, probably not unlike the Noctua paste I have been using all along. I just want to try the REALLY good stuff and see if it REALLY makes a difference!

At any rate, as I try to air cool a G5 Quad, probably the hottest CPU Apple ever brought to market, I need every advantage I can get!
 
Last edited:
  • Like
Reactions: Turbohydramatic
Dang, Kryonaut has a conductivity rating of 12.5. That's really good vs the 8's of the stuff I have around. I've read that TG in an effort to maximize efficiency (which they have) has sacrificed longevity meaning it wears faster. I have no idea if that is true of course or just naysayers being stupid. Still I am intrigued enough to pick up a gram the next time I pull apart and repaste my dualcore 2ghz a1117 cpus and the U3. Last time I repasted this mac, I skipped the U3, so I need to do it next time around.
 
Kryonaut is really aimed at the gamer/overclocker market, where CPUs and GPUs are pegged out for large periods of time. It has the consistency of clay and it seemed to dry rapidly in my experience, clumping up on the CPU and sticking to the paste spreader making it a real pain to work with. As I said earlier, I wouldn't want to use on a largish die as you need to move it quickly if you want to get an even spread under the heatsink. Real world tests (plenty on YT) don't show a night and day difference with much cheaper decent pastes despite the impressive looking W/mK figures. Good luck.
 
All good points @weckart, thanks for the input. I have read enough similar things that I was planning to do a practice run on a spare copper shim first. If that shows that Kryonaut Extreme is too difficult to use to get an even coating (and I will need two such coatings: one on the CPU and one on the shim) then I will drop back to something easier to use.

...or perhaps I will use Kryonaut Extreme on the CPU, as it has a relatively small surface area, and then something that spreads more simply on the shim, which has a significantly larger surface area. I will experiment first and decide once I have done so.

The YT pointer is a good one, thanks. I will pull up a few videos about Kryonaut Extreme and watch them. That will help me to better understand what I am up against.
 
mac57mac57 said:
All good points @weckart, thanks for the input. I have read enough similar things that I was planning to do a practice run on a spare copper shim first. If that shows that Kryonaut Extreme is too difficult to use to get an even coating (and I will need two such coatings: one on the CPU and one on the shim) then I will drop back to something easier to use.

...or perhaps I will use Kryonaut Extreme on the CPU, as it has a relatively small surface area, and then something that spreads more simply on the shim, which has a significantly larger surface area. I will experiment first and decide once I have done so.

The YT pointer is a good one, thanks. I will pull up a few videos about Kryonaut Extreme and watch them. That will help me to better understand what I am up against.
Click to expand...
That's my concern too, I hate thermal paste that isn't easy to apply. I remember I tried mx5 when it came out, I think they had to take it off the market because it was kinda bad lol. I remember having so much trouble with it I just went back to 4. I think they have Mx6 now, But I'm not really in a rush to try it :). It never occurred to me that the pastes thicker and less runny properties had something to do with being higher performance. The tid bit about longevity is interesting. I definitely never want to have to do mine again. at least not for a looong time. Thermal pads can be nice for that reason, but I often I prefer past just cause when it's not an absolute nightmare to replace, you don't have to worry about the thickness of the pads and stuff and its cheap and easy to do so. I usually like to take note of what old paste looks like on PCs I open that are probably a decade plus old, some are pretty terrible, some still somewhat okay but definitely still needing fresh paste.
 
weckart said:
Real world tests (plenty on YT)
Click to expand...

I did in fact pull up several Kryonaut Extreme YT videos and watched them. At least one of the ones I watched reported a 5 C improvement in CPU temps, which is an excellent result.

What really caught my attention however was that the the Kryonaut Extreme paste doesn't seem nearly as hard to spread as people make it out to be. I will know better when I try it myself, but on the videos it looked like a pretty ordinary pasting effort, with the exception of the fact that the presenter used a bit more paste (maybe 50% more) than usual to ensure smooth application. The only down side they noted of this was that it wasted a bit of rather pricey paste!

They also explicitly poo-hoo'd the pasting technique that has the person doing it simply put several drops of paste on different parts of the CPU face, in an "X" or a "P" pattern, and then relying on the screw down pressure of the screws attaching the heat sink on top of it to spread the paste out. This was explicitly rejected as creating areas on the chip face that don't get any paste, creating "hot spots" on the chip that can still lead to thermal shutdown. SO... spread the paste evenly, covering the entire chip face, and all will be well... or at least better than if the dot technique is used.

I feel better about the likely results with Kryonaut Extreme after watching these videos - thanks!
 
weckart said:
Kryonaut is really aimed at the gamer/overclocker market, where CPUs and GPUs are pegged out for large periods of time. It has the consistency of clay and it seemed to dry rapidly in my experience, clumping up on the CPU and sticking to the paste spreader making it a real pain to work with. As I said earlier, I wouldn't want to use on a largish die as you need to move it quickly if you want to get an even spread under the heatsink. Real world tests (plenty on YT) don't show a night and day difference with much cheaper decent pastes despite the impressive looking W/mK figures. Good luck.
Click to expand...
This makes great sense. My use case for this mac is certainly not gaming or OCing it. Truthfully, I am interested in longevity because breaking one of my PMs down is not my idea of a fun afternoon. I have better more pressing things to do so when I do commit to something like that, I want the longest lasting option. It is a difficult metric to find IMO.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back