News M3 MacBook Air hits eye-popping 114 degrees Celsius in stress test and didn't melt — temperature settles down at 100 degrees after thermal throttling

[Admin]

Apple's latest M3-powered MacBook Air has a little heat problem and can easily hit over 100 degrees Celsius in prolonged, intensive workloads.

M3 MacBook Air hits eye-popping 114 degrees Celsius in stress test and didn't melt — temperature settles down at 100 degrees after thermal throttling : Read more

 

[hotaru251]

100 degrees, which is still hot but is still considered safe by the industry.

makes me wonder how the longevity of these will be as heat is bad enough but almsot worse is the high heat to cool cycling over time.

 

[ezst036]

Well there goes the promise of ARM/RISC being smaller thus being cooler and using less power, etc etc.

 

[kremenny]

And the GPU/memory is on the same die. They better make it convenient to swap out the whole circuit board.

 

[kaalus]

What is more noticeable: 20% reduction in performance or 50dB whine from the cooling fans? I'll take the drop in performance any day.

 

[thestryker]

Ever since I saw mods for the M2 MBA I just assumed this is partially a segmentation thing. They could make the cooling on the MBA much better than it is, but that would add weight/size/noise and also put it in the same performance range as the lower end MBP.

 

[hotaru251]

Well there goes the promise of ARM/RISC being smaller thus being cooler and using less power, etc etc.

its apple.

they aren't usually known for cooling solutions.

they try to keep it small as heck. an actual proper thermal solution would likely tame this by a large amount.

 

[George³]

Silicon

Melting point	1687 K (1414 °C, 2577 °F)

From Wikipedia.
What expected to be melting in MacBook when only hotpoint going to 114°?

 

[Neilbob]

Just another example of Apple shooting for form over function. But it's okay, cos Apple get a free pass from their devout following.

It could form the backbone of their next marketing campaign: Our new doobrey has the magical potential to imbue your fingertips with blisters.

I can see people queueing around the block already, credit cards and iPhones, poised and ready for another outlay of multiple thousands, glinting in the early morning sunlight.

 

[cyrusfox]

Silicon
Melting point 1687 K (1414 °C, 2577 °F)
From Wikipedia.
What expected to be melting in MacBook when only hotpoint going to 114°C?

Its not silicon melting that has ever been the issue, its typically Electron migration usually via Copper migration path that leads to shorts and ultimately chip failure. Higher heat will accelerate chip failure at both Silicon and PCB.

Quality assurance on chips uses heat to accelerate chip failure. A tester will repeated burn chips (put in an oven, then run them in a normal or max temp environment) this is to imitate aging, Usually running them through burn in cycles till failure or measure performance degradation and estimate lifetime failure rates based on the curve of a given electrical degradation. Fact is heat(and/or high voltage) will eventually kill electronic circuitry, whether that is 1 year, 5 years, or 100 years is a big deal.

As the macbook only has a 1 year warranty, maybe this is by design, as I think even at 114°C it will pass the warranty threshold, but I would not expect serviceable life up to 5 years... They have other issues that prevent longevity though that may prove to fail sooner [Display, NAND, serialized hall effect sensors, T2 Security Chip, etc...]

 

[Alvar "Miles" Udell]

Product segmentation. The Air isn't intended to be used in the same kind of tasks that the Pro is, else every review would end with "Save $500, skip the Pro, and get the Air instead".

 

[Alpha_Lyrae]

Very likely the junction temperature, so chip edge was around 10-15C cooler or 99-104C. If 114C was chip edge, there'd be some seriously concerning junction temperatures and you'd very likely see instability or full-on overheat protection screen.

Still, Apple is running their silicon to nearly the maximum limit of 115C for logic silicon. This does cause degradation, but since Apple's M3 doesn't clock high, it's not a huge concern. It can probably run this way and still last up to 4-5 years (barring no solder ball failures or any other ancillaries like VRM and such) with adaptive VFS that compares running logic transistors with reference transistor logic.

Toward the end of projected silicon life, it will likely not reach maximum boost of 4.05GHz due to the effects of electromigration and will use more power to achieve lower clocks and reduced performance. This is a design choice, as many of these products will likely be recycled after 5 years, with a vast majority being traded-in after only 2 years.

 

[brandonjclark]

Product segmentation. The Air isn't intended to be used in the same kind of tasks that the Pro is, else every review would end with "Save $500, skip the Pro, and get the Air instead".

And this is why the Air sells so well.

 

[salgado18]

Product segmentation. The Air isn't intended to be used in the same kind of tasks that the Pro is, else every review would end with "Save $500, skip the Pro, and get the Air instead".

This is a bad guy move, or way too naive. The Air is cheaper than the pro, and it's not written anywhere "don't push it to the limit too often". And by the way, why is a product not allowed or recommended to be run at its limit? Isn't the limit a safe spot? Is the Air a factory-overclocked version? Shouldn't it have a lower clock to last longer and work safer?

 

[jmcgaw]

Put the kettle on...

 

[TechLurker]

its apple.

they aren't usually known for cooling solutions.

they try to keep it small as heck. an actual proper thermal solution would likely tame this by a large amount.

Wonder how long until they get AirJet Mini solid state fans into their devices. This is one of those ideal use cases, even if all it does is just blow air over minimally a metal slab acting like a heatsink.