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

XyroTM

macrumors newbie
Original poster
Apr 9, 2021
8
2
I just recently bought a 2017 12 inch macbook and it should get here in a couple of days. ive been seeing apples lackluster cooling solution for the core m3 and i was just wondering if changing the stock "thermal paste" (more like thermal adhesive) for a good high quality thermal pad (like this: Amazon.com: ARCTIC - Thermal Pad 120 x 20 x 1.0 mm Thermal Compound for All Coolers Efficient Thermal Conductivity Gap Filler Safe Handling Easy to Apply: Computers & Accessories) could help the thermals.
 
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
i should also mention that the thermal pads that i have are 1.5 mm thick
 
The pads are definitely a viable alternative to paste. The more important question is why are you considering doing anything with the thermal transfer medium on the CPU?
You don't mention that there is any symptom, and I'm a little bit old school I'm afraid, where I don't fix things unless they're broken.

What is the factory specified thickness of a thermal pad supposed to be? For the cooling mechanism that you have, is 1.5 mm suggested, or 0.5 mm? It's unclear, but if you were asking if one will fit where the other is, most modern cooling assemblies are spring mounted, and will take up any slack providing the cooling transfer medium allows the threads to engage. If you don't know why there are different thicknesses, then you should do some research to determine which thickness will work best for your current situation.

Should we assume that you've already got a certain level of cooling issue to resolve?

P.S. speaking generically, and not about Apple or anything else, taking a computer apart, especially down to the smaller components, stresses the motherboard and the components. This is similar to how extreme heat and extreme cold cycling back and forth over time can cause problems on multi-layer PCBs. Physically cleaning a motherboard, it's components and peripherals does wonders for the longevity of individual components and overall cooling per the original design. If you have benchmarked temperatures, and have noted them going into excess levels over time, going out of spec that is, there are other things that can also cause this to happen.
 
The pads are definitely a viable alternative to paste. The more important question is why are you considering doing anything with the thermal transfer medium on the CPU?
You don't mention that there is any symptom, and I'm a little bit old school I'm afraid, where I don't fix things unless they're broken.

What is the factory specified thickness of a thermal pad supposed to be? For the cooling mechanism that you have, is 1.5 mm suggested, or 0.5 mm? It's unclear, but if you were asking if one will fit where the other is, most modern cooling assemblies are spring mounted, and will take up any slack providing the cooling transfer medium allows the threads to engage. If you don't know why there are different thicknesses, then you should do some research to determine which thickness will work best for your current situation.

Should we assume that you've already got a certain level of cooling issue to resolve?

P.S. speaking generically, and not about Apple or anything else, taking a computer apart, especially down to the smaller components, stresses the motherboard and the components. This is similar to how extreme heat and extreme cold cycling back and forth over time can cause problems on multi-layer PCBs. Physically cleaning a motherboard, it's components and peripherals does wonders for the longevity of individual components and overall cooling per the original design. If you have benchmarked temperatures, and have noted them going into excess levels over time, going out of spec that is, there are other things that can also cause this to happen.
the one problem that im trying to solve is excess heat as these computers are kind of know for getting so hot and cooling down so fast that the board starts to flex and the cpu solder points get ripped off. so by replacing the thermal paste for something better it should reduce the heat output therefore making the computer quite a bit more reliable
 
  • Like
Reactions: bizthewiz
Oh! Thanks for the context, I couldn't tell if you actually had a problem yet or were anticipating one.

Here is the advice that you probably don't want to hear. You should not change anything on the machine, especially opening it up and taking it apart, until your warranty is up.

Apple is notorious for finding ways to not have to pay to repair their devices. If they suspect anything in the way of you being in that machine, they will flat out reject doing anything for free. Those genius texts and AASPs don't know how to do anything except take the covers off, look for signs of the customer tampering with the device, and look at the moisture sensors. Apple does not repair devices. #RTR

If you have a heating problem with this unit, it will be noticeable immediately. You won't have to wait until after a warranty. Where they will fix everything for free to determine if you actually have a problem or not. I strongly suggest to all my customers that they utilize any warranties that they have paid for before shelling out any money.

The advice, and call it an opinion, that I give you is that unless you want to void your warranty day one, which is your choice, I would wait and document any kind of thermal problems that you might have with that particular unit, an exhaust any free services, repairs, or upgrades that you can get out of Apple before taking that cooling assembly apart.

Apple will not sympathize with your actions to avoid overheating, and like a good insurance company, will use it as an excuse not to pay you what you are due.

I'm sure someone else here will give you advice more like what you want to hear, but this is a brand new unit, and that is my opinion in trying to help you.
 
  • Like
Reactions: XyroTM
the one problem that im trying to solve is excess heat as these computers are kind of know for getting so hot and cooling down so fast that the board starts to flex and the cpu solder points get ripped off. so by replacing the thermal paste for something better it should reduce the heat output therefore making the computer quite a bit more reliable
I have 2x 12" rMBs. From my experience, the heat transfer bottleneck is not die to heatsink (which is where the thermal paste is). The bottleneck is aluminum case to ambient air. In other words, if you care about avoiding the hot/cool cycles: (1) it's probably better to just use the laptop where there is good ambient airflow, or (2) this isn't the best laptop for you.

Basically, from what I can tell, the CPU die to heatsink/bracket transfer is fine and the heatsink/bracket to case transfer is fine. You can tell because the case gets hot a hell under load - sometimes too hot to even have on your lap. The issue is the case stays hot other than slow dissipation into the ambient air.

I don't feel like searching; someone else in this forums did a bunch of benchmarking and temperature tests a while back. One of the tests showed that the Macbook performed materially better and stayed significantly cooler on a granite kitchen counter as compared to on a wooden desk. This is because granite is a good heat conductor and is good at drawing the heat away from the aluminum case, whereas wood is the opposite and acts more as a thermal insulator than a heat conductor.

So that's our thermal bottleneck: Cooling the aluminum case. It's certainly not pretty, but I bet something like this pointed back at the body would do a lot more than replacing thermal pads.
 
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
I have 2x 12" rMBs. From my experience, the heat transfer bottleneck is not die to heatsink (which is where the thermal paste is). The bottleneck is aluminum case to ambient air. In other words, if you care about avoiding the hot/cool cycles: (1) it's probably better to just use the laptop where there is good ambient airflow, or (2) this isn't the best laptop for you.

Basically, from what I can tell, the CPU die to heatsink/bracket transfer is fine and the heatsink/bracket to case transfer is fine. You can tell because under load, the case gets hot a hell - sometimes too hot to even have on your lap. The issue is other than slow dissipation into ambient air, the case stays hot.

I don't feel like searching, but someone else in this forums did a bunch of benchmarking and temperature tests a while back. One of the tests showed that the Macbook performed materially better and stayed significantly cooler on a granite kitchen counter as compared to on a wooden desk. This is because granite is a good heat conductor and is good at drawing the heat away from the aluminum case, whereas wood is the opposite and acts more as a thermal insulator than a heat conductor.

So that's our thermal bottleneck: Cooling the aluminum case. It's certainly not pretty, but I bet something like this pointed back at the body would do a lot more than replacing thermal pads.
ahhhh yes i saw that test too (this: This is how the 2017 MacBook Core m3, i5, i7 behave during sustained CPU load: 10+ runs of Cinebench | MacRumors Forums) and really im just gonna be using this laptop for school. so i will mostly be editing word documents, connecting via microsoft teams, listening to music, a couple of edge tabs open (yes i use edge, its not that bad of a browser anymore and it actually performs noticeably better on my 2015 macbook air 13") so i think it should be fine. im not planning on using blender on this laptop any time soon, thats why i have a ryzen 5 2600 LOL
 
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
I have 2x 12" rMBs. From my experience, the heat transfer bottleneck is not die to heatsink (which is where the thermal paste is). The bottleneck is aluminum case to ambient air. In other words, if you care about avoiding the hot/cool cycles: (1) it's probably better to just use the laptop where there is good ambient airflow, or (2) this isn't the best laptop for you.

Basically, from what I can tell, the CPU die to heatsink/bracket transfer is fine and the heatsink/bracket to case transfer is fine. You can tell because the case gets hot a hell under load - sometimes too hot to even have on your lap. The issue is the case stays hot other than slow dissipation into the ambient air.

I don't feel like searching; someone else in this forums did a bunch of benchmarking and temperature tests a while back. One of the tests showed that the Macbook performed materially better and stayed significantly cooler on a granite kitchen counter as compared to on a wooden desk. This is because granite is a good heat conductor and is good at drawing the heat away from the aluminum case, whereas wood is the opposite and acts more as a thermal insulator than a heat conductor.

So that's our thermal bottleneck: Cooling the aluminum case. It's certainly not pretty, but I bet something like this pointed back at the body would do a lot more than replacing thermal pads.
also i bet putting the laptop on a laptop stand would help things a lot as well
 
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
I don't feel like searching; someone else in this forums did a bunch of benchmarking and temperature tests a while back. One of the tests showed that the Macbook performed materially better and stayed significantly cooler on a granite kitchen counter as compared to on a wooden desk. This is because granite is a good heat conductor and is good at drawing the heat away from the aluminum case, whereas wood is the opposite and acts more as a thermal insulator than a heat conductor.
Glad to be of service. :)

Anyhow, my 2017 Core m3 MacBook is fine so far. No perceivable difference in thermal performance compared to when I bought it four years ago. That said, I haven't used it much in the past year since I'm stuck at home and use an iMac at home.

So, I'm with the others that say don't fix it unless it's broken. So far the only Macs I've bothered replacing the thermal pads was an old 2008 white MacBook that was overheating (partially due to dust clogging the fan and partially due to 10 year-old thermal pads), and a 2006 Mac Pro when I was changing the 14 year-old CPUs anyway. However, I replaced the pads with thermal paste. These setups allowed thermal paste to be used. IMO properly applied good quality thermal paste is usually quite a bit better than thermal pads, but only if the heatsink mount allows proper contact for thermal paste. This is not always the case in designs using thermal pads. But like I said, don't bother unless you need to.
 
I have 2 rMB: 2015 m3 and 2017 i5 - both are stock.

No heating issues - runs a lot cooler than my fan powered Intel: Air, Pro, Mini, iMac...

For a passive cooling device in a ultra thin form, it runs very cool.

Even after 6 years, the 2015 rMB runs like a champ, although the battery is at 75% charge and reports need service.
 
So far the only Macs I've bothered replacing the thermal pads was an old 2008 white MacBook that was overheating (partially due to dust clogging the fan and partially due to 10 year-old thermal pads), and a 2006 Mac Pro when I was changing the 14 year-old CPUs anyway. However, I replaced the pads with thermal paste. These setups allowed thermal paste to be used. IMO properly applied good quality thermal paste is usually quite a bit better than thermal pads, but only if the heatsink mount allows proper contact for thermal paste. This is not always the case in designs using thermal pads. But like I said, don't bother unless you need to.
I agree except on the rMB. It just doesn't seem like the thermal paste on these machines is the thermal bottleneck that needs to be addressed. I have not tried, but my guess is properly applied good quality thermal paste would actually have no effect on the rMB.
 
I agree except on the rMB. It just doesn't seem like the thermal paste on these machines is the thermal bottleneck that needs to be addressed. I have not tried, but my guess is properly applied good quality thermal paste would actually have no effect on the rMB.
Well that's the thing. I haven't needed it on the rMB, and my advice is don't bother redoing the thermal paste if you don't need to do it.

That said, if you had to remove that sink for some reason, you'll want to replace the thermal paste anyway, and it may be possible that it could help reduce heat related thermal throttling. My granite counter test does not "prove" that computer to outside heat transfer is the main bottleneck. All that indicates is that it is one of the factors.
 
Well this post begs to differ regarding this computer:


Clearly (and I've noticed it since the beginning on my 2017 A1534) the CPU shoots to 99 degrees C and stays there under a sustained load, this is the definition of throttling. Sure the bottom gets warm, but the warmer we can make it get the less throttling and better perf we can get out of it.

I'm willing to take it apart (carefully) at this point now that it is worth only around $500 on the aftermarket, since I'd like for this computer to be able to consistently deliver reasonable performance and it is still more compact than any other solution that has a keyboard. My iPad Pro with magic keyboard feels twice as hefty. Probably because it is.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.