Hover over user name, click Ignore, for anyone wondering.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
M2 overheating. No solutions so far?
- Thread starter sunny5
- Start date
- Sort by reaction score
Telling a fact hurts you I guess?
Do you mean underestimated? Also, doubtful.
What does that mean? M2 outperforms the M1 at the same wattage and the new MacBook Air chassis can dissipate the same amount of sustained heat output as the old one. There is no evidence whatsoever that it's not working as intended.
Any link to support your claim? Unfortunately, NotebookCheck's M2 MBA review is not as complete as the M1 MBA review.
Dissipating the same amount of sustained heat may not be enough. The 13" MBP M2 gets hotter than the 13" MBP M1.
M2 13" MBP:
Apple MacBook Pro 13 2022 M2 Laptop Review – Debut for the new Apple M2
Notebookcheck reviews the updates 2022 Apple MacBook Pro 13 with the brand-new Apple M2 SoC and 16 GB RAM.
M1 13" MBP:
Apple MacBook Pro 13 2020 Laptop Review: The entry-level Pro also gets the M1 performance boost
Notebookcheck reviews the entry-level MacBook Pro 13 2020 with Apple's new M1 processor. Does the active cooling solution make a noticeable difference?
There were a bunch of results posted here and in other threads showing that M1 Air outperforms M1 Air in sustained operation. what other proof do you need?
You are falling into the same trap as the YouTubers confusing peak and sustained power dissipation. There is ample evidence that M2 can consume more power under load. But combined with the evidence that its sustained performance is improved over M1 it simply means that M2 has a bigger dynamic performance delta. In other words, it has a higher peak performance ceiling than M1.
The fallacy is in assuming that since M2 can “boost harder” it also requires better cooling to provide better sustained performance. But this is not so, as reviews clearly show.
According to iFixit, M2 doesn't even have a heatsink! It only has a thin metal shielding which doesn't do anything! It is worse than M1 MBA for sure. Damn you, Apple.
The M2 gets work done faster. We’re two pages into your thread and all your doing is posting videos that have no technical analysis.
Do better.
He is showing that it doesnt have the same heat sink as before (which may or may not have improved performance). There are plenty of results out there showing the performance. For many including myself its underwhelming and there is plenty of evidence to show this. It doesn’t mean apple did it by mistake or anything else people are interpreting into it. What it does mean is that its not the machine for me in this iteration. I used my m1 air for basic things most the time but liked to fire up some strategy games that run well on mac (single player anyways multi play is another story). I wanted the M2 air but since in a gaming scenario there is not much to be gained (for me) i dont see the point and just went with the 14 inch.
How do you know it? How much heat does the M1 MBA and M2 MBA dissipate?
If the M2 gets hotter and its sustained performance is better than the M1, either the benchmark is not pushing hard enough or Apple is "limiting" the M2.
For instance, a "limited" 6800U performs better than an unrestricted 6800U because it has no thermal throttling.
Last edited:
Between 7 and 10 watts depending in the ambient temperature. I don’t have any links handy unfortunately but that’s what I remember being reported.
The iFixit people might want to update their heat transfer materials knowledge. The metal is just an RFI shield. Most of the heat is distributed by the black pyrolytic graphite sheet covering the metal shield, which contacts the outer case for heat dissipation. PGS has several times the thermal conductivity of copper.
If you need sustained compute performance, get an MBP. If compute loads are bursty, the M2 will outperform the M1. You buy the computer that suits your needs. I run a sustained load on my laptop maybe two days out of the year, so when I replace this one, it will very possibly be with an Air. I don't buy computers to run benchmarks on them. If you do, fine, but don't try to convince me that somehow the M2 as used in the Air is slower than the M1, for appropriate loads.
lol and only morons make mountain out of a molehill
I can't believe this is still going. My roommate has an M2 Air. I've used it. Made it play 4k HDR videos while I web browsed and opened tons of documents and did all the things a normal Air user will do. It doesn't overheat. Didn't even get warm.
Again, the M2 only "overheats" if you make it do things that should be done on a higher-end laptop anyway. And even when it's "overheated," it's still faster than the M1. Get over it already.
Again, the M2 only "overheats" if you make it do things that should be done on a higher-end laptop anyway. And even when it's "overheated," it's still faster than the M1. Get over it already.
Apple has been shielding the bottom case with a sheet of some kind of black material for a very long time; all that's new here is that it seems to be on the heatspreader rather than the bottom case itself. See for example the inner side of a M1 Air bottom case:
This is also present on all MacBook models with fans.
Apple doesn't want a low resistance thermal (or electrical) path between internal components and the metal shell. The computer is designed to be usable on your lap, and even temperatures which aren't immediately painful can cause injuries when the exposure is long enough. (Ever heard about sous vide slow cooking at low temps? Same deal.) There are medical conditions which reduce the ability to feel mild discomfort and pain, there are going to be users who fall asleep with the computer turned on, and so on. You have to protect them, so Apple designs in some insulation to keep surface temps low enough to be very safe, no matter how many forums posters and clickbait youtubers get angry because it limits performance.
With that in mind, folks need to understand that there never was any real reason to believe that M1 and M2 Airs are significantly different in thermal performance. In a fanless computer, the path for heat to leave is the computer's surface, and if the interior thermal components are carefully engineered to limit the rate at which heat energy can enter that surface so it never gets too hot, that's what controls the ultimate power limit after internal components are fully heat soaked. M1 and M2 Airs use the same case material and have almost identical surface area, which are the other important variables here, and guess what? As @leman keeps pointing out, both M1 and M2 Air seem to settle out at about 10W SoC power in long running tests.
So the "big heatsink" inside the M1 Air which people are up in arms about doesn't really matter. At best, it's just delaying the inevitable - it's a slug of metal which, when cold, will absorb some heat energy for a while until it's hot. But since the overall mass of the M1 and M2 Air is so similar, and the area of the metal plate attached to the M2 SoC so much bigger (helping it transfer heat to other internal mass), I doubt it's even all that significant.
Sadly, none of the testing I've seen has really tried to investigate this question in a sound way. What we need is a direct comparison of frequency drop / SoC temperature rise over time between the M1 and M2 Air on identical loads. We know the steady state power is about the same, so the interesting question is which one reaches its steady state faster?
Instead, all we get is youtube reviewers confirmation biasing themselves into believing that the M2 Air must be throttling bad because it doesn't have something which looks heatsinky enough to them, even though what data we do have suggests it's probably just fine. But I'd like to get some more definitive data.
I forgot to mention one subtlety here. From what we know so far, if Apple had built the M2 Air like they did the M2 13" MBP, no change at all in the chassis / thermal design, the M2 version probably would reach steady state faster. I say this because we know that the big power consuming portions of the M2 SoC (such as CPU cores) can reach slightly higher absolute power figures than their M1 counterparts. However, those cores are simultaneously more power efficient - they're getting more computing work done per joule of energy. (Note the different units in play here - watts AKA power are a rate unit, joules are an energy unit. 1 Watt = 1 Joule/second.)
So, even though M2 can draw more instantaneous power from the battery, it's getting more work done per second than M1. That means it's not automatically a terrible thing if M2 Airs reach thermal equilibrium (aka fully throttled) a bit faster than M1 Airs. As long as they still get more work done before then, they're ahead.
People's opinions about this kind of thing are often a reflection of whether they understand basic calculus (derivatives and integrals) and physics.
Dude with this and the previous post you're missing the fact that you are wrong. The internals are different and it does matter. Apple intended it this way sure. But still doesn't change the fact the people were expecting more. Why is there so much interest in putting that down and disregarding it. I wanted more from the machine myself, im not here bitching about it by creating a bunch of posts but is doesn't mean its an irrelevant desire. Anyways not gonna argue about it as that's all I have to say about it.
In regards to the OPs tagline yes there is a solution for M2 over heating. Move to Antartica and do your daily computing outside with a Elon Musk provided Starlink or even better no internet connection and the inability to read forums.
The two sets of data show that the M1 MBP13 and M2 MBP13 have pretty close to the same thermals on their surfaces, under the load they were under.Any link to support your claim? Unfortunately, NotebookCheck's M2 MBA review is not as complete as the M1 MBA review.
Dissipating the same amount of sustained heat may not be enough. The 13" MBP M2 gets hotter than the 13" MBP M1.
M2 13" MBP:
View attachment 2033127
Apple MacBook Pro 13 2022 M2 Laptop Review – Debut for the new Apple M2
Notebookcheck reviews the updates 2022 Apple MacBook Pro 13 with the brand-new Apple M2 SoC and 16 GB RAM.www.notebookcheck.net
M1 13" MBP:
View attachment 2033128
Apple MacBook Pro 13 2020 Laptop Review: The entry-level Pro also gets the M1 performance boost
Notebookcheck reviews the entry-level MacBook Pro 13 2020 with Apple's new M1 processor. Does the active cooling solution make a noticeable difference?www.notebookcheck.net
If you look at the room temperature, the M1 MBP13 had a room temperature of 70 °F while M2 MBP had a room temperature of 73 °F, about 3 °F difference. Actually about 3.4 °F difference if you convert the Celsius numbers. The differences in surface thermals are about 3 to 6 °F. Once you account for the ambient temperature being different, they are looking pretty close. Once you get down to the 1 to 2 °F level differences, your controls and measurements have to get a lot better, like the table that the laptop rested on has to be the same and be at the same heat soak, with no variable sunlight exposure on both the table and laptop, the laptops are at the same heat soak, etc.
Also be mindful that the M2 is doing about 20% to 40% more computations per unit time at those surface temperatures.
Dude I didn't say the internals are the same, learn how to understand points made by discussing hypotheticals we know not to be true.
Unless it turns out that heavy sustained loads cause the M2 Air to reach thermal equilibrium (and thus worst case throttling) significantly quicker than the M1 Air does, I don't think you have any real justification for disappointment. From the information presented in reviews so far, we don't actually know this one way or the other.
I will say that some of the M2 Air benchmarks sure look like the throttling is still quite gradual, just as it was on the M1 Air. But before I'm going to say it's fine, I want to see data from an identical M1 test under identical conditions (ambient temp, cold soak, etc) plotted on the same scale side by side with the M2 data. Then we'd actually have something concrete to discuss!
IMO, you need to start thinking more critically. You're just accepting a lazy narrative that's been pushed on you by a few video reviews. A paraphrase of the usual take is roughly: 'There's no real heatsink, that must mean Apple cheaped out and it's throttling bad, far worse than the M1 Air, and look! these benchmarks show the M2 Air throttling! Riot!' But none of them have taken the step of actually doing a detailed, methodical comparison of throttling behavior between M1 and M2 Air. They just say OMG the M2 Air throttles, and that's it. So how do you even know the new design is worse than the old?
Or are you saying that you expected thermal performance substantially better than the M1 Air chassis? That's not in the cards for any fanless design, so if that's what you wanted you were never going to be happy.
Wrong. I have major reason for disappointment, considering that I wanted it to game occasionally and the performance in that particular workload is hardly an improvement if at all over the m1. And oh yea you got me ill go back and "educate" myself so I can understand points discussing hypotheticals. You said it didn't matter and it does to some of us.
You are doing exactly what you are falsely accusing me of. I do not need to think more critically. I have watched every bit of info and figured it won't work for me. For many it will. I could buy one and waste mine and the store I buy it from time because I know it won't work for me. Do you get that? FOR ME. I am not pushing any agenda and anyone that wants it and can afford it should get it. For me the killer was a video of two machines one with the 8core gpu and 10core and the 10core performed worse. These were the same temps and conditions (side by side in fact) and that was good enough for me to see that the 10core fanless machine I wanted wouldn't do what I wanted. Is that too difficult to understand?
You are doing exactly what you are falsely accusing me of. I do not need to think more critically. I have watched every bit of info and figured it won't work for me. For many it will. I could buy one and waste mine and the store I buy it from time because I know it won't work for me. Do you get that? FOR ME. I am not pushing any agenda and anyone that wants it and can afford it should get it. For me the killer was a video of two machines one with the 8core gpu and 10core and the 10core performed worse. These were the same temps and conditions (side by side in fact) and that was good enough for me to see that the 10core fanless machine I wanted wouldn't do what I wanted. Is that too difficult to understand?
Lack of heat sink already proves that Apple doesn't know how to make a computer. Is this how Mac fanboys support Apple's failure?
Just keep waving those hands. You haven’t even established that here is an issue.