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

M2 fanless heat issue

mr_jomo said:
Yeah - that thing looms massive

Guess that explains the much improved sustained performance over the M1 Air that the Verge documented.
Click to expand...
The usual suspects on YouTube and forum dullards will complain about throttling as if it’s something unexpected but the performance is great and in every single aspect it is better than the M1 Air which was already widely hailed as having excellent performance. The reports of the underside of the chassis getting hot (which also occurred on the M1 Air) are also indicative of the heatsink’s efficiency.

I’m really looking forward to notebookcheck‘s review whereupon they”ll do thermal testing in a somewhat more methodical and thorough manner than any other reviews have so far.
 
Last edited:
  • Like
  • Angry
Reactions: calstanford, jdb8167 and mr_jomo
Chancha said:
Could you point us to these photos? Would like to see how it differs.
Click to expand...
M1 Air

55CD0762-997E-47E8-B994-DDCAB7C55CF6.jpeg


M2 Air

10380505-45B3-4750-A30A-76CE8FADAE2B.jpeg
 
Chancha said:
I wonder if these reviewers are under NDA constrains to not further remove the plastic cover? From this shot it is hard to guess where exactly the M2 chip is; if it is near center with the four screw holes, wouldn't that trackpad cable directly underneath be a problem?
Click to expand...
Absolutely they’ll be under NDA not to open it up whereas independent reviews with purchased units will have no such constraints.

As for the trackpad cable that is seemingly an odd design choice. Obviously it’ll have a thermal rating and Apple must be certain that this will not be exceeded otherwise they wouldn’t have routed it there. But yeah it does look somewhat incongruous.

There are more photos here - https://9to5mac.com/2022/07/13/exclusive-m2-macbook-air-internals-in-hands-on-photos/
 
And yet it is again MaxTech who puts out the earliest full teardown video for the M2 MBA. It is now confirmed that the logic board spans only 2/3 of the space with 1/3 being passive heat sink, NAND slot is one empty for base 256GB as expected. The heat paste is possible to be repasted with some effort after disassembly but not overly difficult.
 
Yeah just watched the Maxtech teardown video. Seems that I was wrong and the heatsink doesn’t have the mass that I thought it did from just looking at the photo’s. A shame but still it‘s always faster computationally than the M1 Air and indeed even when heavily throttling it generally drops down to about the same performance as or just slightly below the performance that the M1 Air has without throttling, for both the CPU and the GPU.
 
  • Like
Reactions: yitwail
After a day of using the Air, on the couple of times where the CPU & GPU cores have been working, the machine warms up a bit (nothing uncomfortable) and cools back down again very quickly.
 
  • Like
Reactions: yitwail, MayaUser and daavee80
ascender said:
After a day of using the Air, on the couple of times where the CPU & GPU cores have been working, the machine warms up a bit (nothing uncomfortable) and cools back down again very quickly.
Click to expand...
Glad to hear it. Seems like the cooling system is doing it’s job well and just as effectively as the M1 Air which is widely regarded as having decent thermals. Throttling and heat are to be expected with a passively cooled system and unfortunately hyperbolic, clickbait articles and videos about thermal issues are to be expected also.
 
  • Like
Reactions: dustin_ and ascender
ascender said:
After a day of using the Air, on the couple of times where the CPU & GPU cores have been working, the machine warms up a bit (nothing uncomfortable) and cools back down again very quickly.
Click to expand...
i can confirm that, after a day in the sun on my backyard the system reached around 104C in my projects (here are around 34C, even more in direct sunlight) the bottom and the part above the keyboard gets hot, but after 3-4 minutes the system gets cold very quickly and the system is under 90C..so no matter if you get some throttling the system is thought out to be used in comfort ways even if that means loosing power. My 16" Intel Mbp was hot all the time, even when it was throttling
 
  • Like
Reactions: dustin_, daavee80, ascender and 1 other person
MayaUser said:
i can confirm that, after a day in the sun on my backyard the system reached around 104C in my projects (here are around 34C, even more in direct sunlight) the bottom and the part above the keyboard gets hot, but after 3-4 minutes the system gets cold very quickly and the system is under 90C..so no matter if you get some throttling the system is thought out to be used in comfort ways even if that means loosing power. My 16" Intel Mbp was hot all the time, even when it was throttling
Click to expand...
That’s actually pretty impressive. Apologies if you already know this but I would be careful using it in direct sunlight at those sorts of environmental temperatures though. Apple‘s user guidelines state that operating temperatures should be no higher than 35C and while you won’t damage any of the electronic components such as the SoC or PCB in that sort of heat it really won’t be good for the health of the batteries.
 
I have a 10/16/1TB and it barely gets warm. I've been downloading and installing. Thought it would get hotter. It's rather warm in the room also.

Maybe it will? I'm sure it would if I were to push it. I have no idea how this thing throttles. Does it throttle normally under light loads, so that's why it isn't heating up? That would make sense. No use in pushing an idle machine.
 
  • Like
Reactions: mr_jomo
For some context, it appears the CineBench score after thirty minutes of repeated runs on the M2 still exceeds the CineBench score of a cold first-run M1.

Based on Luke Miani’s test and some Googling.
 
  • Like
Reactions: daavee80
Even if you find Max Tech to be over-hyperbolic or not, the data from the channel is always useful when you take your time to make up your own mind about it.

For instance: the latest "OMG the M2 throttles like ****" video showed that:
(A) Even fully heat saturated and throttled the M2 Air performs about 12% better than an M1 in the same scenario, and
(B) If you go back and dig in the M1 Air throttling video, you'll find that the M2 Air delivers the throttled performance increase at essentially the same wattage as the throttled M1 - lending credence to Apple's claim on improve performance pr. watt in the M2 vs the M1.

All good stuff IMHO.
 
  • Like
  • Love
Reactions: Wizec and Love-hate 🍏 relationship
daavee80 said:
Absolutely they’ll be under NDA not to open it up whereas independent reviews with purchased units will have no such constraints.

As for the trackpad cable that is seemingly an odd design choice. Obviously it’ll have a thermal rating and Apple must be certain that this will not be exceeded otherwise they wouldn’t have routed it there. But yeah it does look somewhat incongruous.

There are more photos here - https://9to5mac.com/2022/07/13/exclusive-m2-macbook-air-internals-in-hands-on-photos/
Click to expand...
what is the problem with trackpad? heat?
 
0089294 said:
what is the problem with trackpad? heat?
Click to expand...
No problem really. More my incorrect assumptions than anything. If you look at the internals of the M2 Air you can see that the trackpad ribbon cable runs over what I thought was the surface of the heatsink. In actuality it’s well insulated and while it’ll warm up under load (along with everything else) from what I can see there will not be any heat-related issues caused by this design.
 
  • Like
Reactions: Love-hate 🍏 relationship
Looking at the latest m2 Air stress test videos/benchmarks even under worst case scenario benchmarking tests the drop in performance is only around 25% tops. So worst case scenario even completely heat soaked after running benchmarks for 15 minutes solid it's still going to be a quick laptop even at 75% of it's geekbench/etc scores. In most real world apps/usage I suspect the difference wll be far less.
 
hungryghosty said:
Looking at the latest m2 Air stress test videos/benchmarks even under worst case scenario benchmarking tests the drop in performance is only around 25% tops. So worst case scenario even completely heat soaked after running benchmarks for 15 minutes solid it's still going to be a quick laptop even at 75% of it's geekbench/etc scores. In most real world apps/usage I suspect the difference wll be far less.
Click to expand...
It’s incredible really. Under very heavy load and throttling the performance is in line with a 2019 i9 MacBook Pro 16 5500M when that machine is under a light load! Think about how much faster the M2 Air is when the i9 machine is heat soaked... In every respect the M2 Air is faster outside of a few bits of x86 optimised software that have to run through Rosetta.
All in a thin and light 13 inch passively cooled ultrabook with 15 hour battery life. And yet this forum is swamped with people saying that it’s underpowered or too slow when throttling or ‘crippled’ or only suited to web browsing and office work and for any heavy lifting one should buy a Pro. The lack of critical thinking is crazy, although sadly to be expected.
Performance-wise the M2 Air is an absolute monster.
 
Last edited:
  • Like
Reactions: JFR0317 and RCElectricFlyer
Stated in another thread - my M2 is cooler than my old 2017 MacBook Air - doing basically the same things for 2 days
 
  • Like
Reactions: mr_jomo
daavee80 said:
No problem really. More my incorrect assumptions than anything. If you look at the internals of the M2 Air you can see that the trackpad ribbon cable runs over what I thought was the surface of the heatsink. In actuality it’s well insulated and while it’ll warm up under load (along with everything else) from what I can see there will not be any heat-related issues caused by this design.
Click to expand...
But it is still an odd choice to position the trackpad cable crossing exactly where the M2 chip "cover" is. Even if insulation is sufficient, there are cooler places on the board for that cable to cross, so at the end this is creating unnecessary potential heat build up into the cable. We are going to get heatwaves soon so every degree counts.
 
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