New MacBook Air has useless fan?

[amyk]

Your thoughts?

[wraper]

It might create some airflow over that heatsink if enclosure creates sort of air guide from the heatsink to the fan. However even if it's made so, it's a very crappy cooling nonetheless. No way it can even remotely perform as a heatpipe and proper heatsink behind the blower fan.

[Someone]

Not a new discovery or a new design:
https://www.ifixit.com/Teardown/MacBook+Air+13-Inch+Retina+2018+Teardown/115201
Cooling a 7W ultra low power CPU with integrated graphics is a little different to the hardware required for most laptops.

[BravoV]

Suddenly the words "planned obsolescence" pop out in my mind. 

[PartialDischarge]

Whenever Louis I'm the man Rossman goes from repairman to design engineer he is just embarrassing himself, covered by the suppport of a herd of fans that will cheer whatever he says against apple.

I think the ideas was to use the cavity from the convex shape of the clamshell chassis as an air current rectifier, and suck air from one side, through the fins, to the other side where the fan's inlet plane is located, then blow the air through the outlet port.

That is how it works

I'm pretty sure Apple did the simulation, and it works

I'm pretty sure they placed temp probes inside

However even if it's made so, it's a very crappy cooling nonetheless. No way it can even remotely perform as a heatpipe and proper heatsink behind the blower fan.

It is not a crappy design based on your (or my) impression. It will perform or not based on real data, steady state and transient temperature profiles.
Judging a cooling system by its looks its worse than judging a circuit by its looks, you need measurements.

[BrianHG]

It is not a crappy design based on your (or my) impression. It will perform or not based on real data, steady state and transient temperature profiles.
Judging a cooling system by its looks its worse than judging a circuit by its looks, you need measurements.

But still, wouldn't a properly piped fan design mean the fan would not have to work as hard?
More efficient meaning a slower fan, saving battery power and lower RPM means lower noise...

[Cyberdragon]

But still, wouldn't a properly piped fan design mean the fan would not have to work as hard?
More efficient meaning a slower fan, saving battery power and lower RPM means lower noise...

The CPU on the MacBook and MacBook Air is supposed to only operate when you open a program. Otherwise, it is supposed to stay idle.
That's why we have hardware video decoder, hardware encryption coprocessor and hardware {put_function_in_there}.

It is not a number crunching machine. If you need a transcoding warrior or a simulation powerhouse, you are buying the wrong computer.

That is still poor design then if the cooling is the limiting factor of PC performance. That's like shipping a car with a tiny radiator and saying "don't expect to rally with it".

[Someone]

It is not a crappy design based on your (or my) impression. It will perform or not based on real data, steady state and transient temperature profiles.
Judging a cooling system by its looks its worse than judging a circuit by its looks, you need measurements.

But still, wouldn't a properly piped fan design mean the fan would not have to work as hard?
More efficient meaning a slower fan, saving battery power and lower RPM means lower noise...

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

That is still poor design then if the cooling is the limiting factor of PC performance.

If you can find a single thin and light high performance computer (gaming laptop with dGPU and i7/i9-HQ/HK under 20mm thickness) without thermal throttling during CPU+GPU dual stress testing, I'd like to hear about it.

Even oversized gaming "laptops" from Alienware etc hit thermal limits on many of their configurations.

[legacy]

That's like shipping a car with a tiny radiator and saying "don't expect to rally with it".

Indeed! Facts are: my brother's Mac Pro got seriously damaged for overheating during an FEM analysis and repaired TWO times before it got replaced with a serious laptop.

[maginnovision]

My HP gaming laptop doesn't throttle... It's also 35mm thick and quite hefty. When I installed the second drive the cooling appeared to be adequate but it is a little more of a dust trap than I'd like.

[langwadt]

But still, wouldn't a properly piped fan design mean the fan would not have to work as hard?
More efficient meaning a slower fan, saving battery power and lower RPM means lower noise...

The CPU on the MacBook and MacBook Air is supposed to only operate when you open a program. Otherwise, it is supposed to stay idle.
That's why we have hardware video decoder, hardware encryption coprocessor and hardware {put_function_in_there}.

It is not a number crunching machine. If you need a transcoding warrior or a simulation powerhouse, you are buying the wrong computer.

That is still poor design then if the cooling is the limiting factor of PC performance. That's like shipping a car with a tiny radiator and saying "don't expect to rally with it".

most cars will overheat if you take them on a racetrack and drive in anger, cooling system is insufficient for the high power of constant acceleration and braking keeping the average speed down

[Kilrah]

If you can find a single thin and light high performance computer (gaming laptop with dGPU and i7/i9-HQ/HK under 20mm thickness) without thermal throttling during CPU+GPU dual stress testing, I'd like to hear about it.

That's irrelevant. If something thermal throttles because the cooling solution when it would have been easy to do better without drawbacks then it's stupid design.
My Retina macbook throttles, but in this case the design intent was for it to be fanless so it's normal. But IF that macbook air throttles because of the colling design when a heatpipe could have prevented it then it's poor design since you're not losing something by putting one.

Lots of ifs since we don't know if it does, but then again we can expect it to since pretty much every Mac does.

[wraper]

I'm pretty sure Apple did the simulation, and it works

I'm pretty sure they placed temp probes inside

I'm pretty sure Apple made warranty extension programs (of which buyers usually are not aware of or on terms most of them will miss it) multiple times because of chip dying as a consequence of overheating.

[wraper]

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

Adding heatpipe with a heatsink behind the blower fan does nothing to prevent fan from moving air across the PCB from intake side .

[SiliconWizard]

If you can find a single thin and light high performance computer (gaming laptop with dGPU and i7/i9-HQ/HK under 20mm thickness) without thermal throttling during CPU+GPU dual stress testing, I'd like to hear about it.

That's irrelevant. If something thermal throttles because the cooling solution when it would have been easy to do better without drawbacks then it's stupid design.

You're welcome to design better within Apple's constraints (size, weight, noise) and submit that to them - I'm sure they'd be happy to hire you.

Performance laptops that can run for extended periods of time at full clock speed are usually very chunky and heavy. They often look more like "portable computers" than thin laptops.

[Kilrah]

You're welcome to design better within Apple's constraints (size, weight, noise) and submit that to them

Adding a heatpipe in the free space we see would not make more noise, make the machine any bigger or weigh significantly more.

- I'm sure they'd be happy to hire you.

The constraints they set for themselves and the compromises that result of it are not ones I would agree to work towards.

[Someone]

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

Adding heatpipe with a heatsink behind the blower fan does nothing to prevent fan from moving air across the PCB from intake side .

There isn't space for a heat pipe and associated heatsink in that layout. As mentioned above you'd be radically redesigning the entire system to fit that sort of design in, and there is unlikely the space to fit heat pipes without impacting or compromising something else. If you're convinced its so easy, make a cooling package for some of these laptops and sell it to all these people who want better cooling, [sarcasm]it'll be a huge market and you'll be rich[\sarcasm].

[Someone]

You're welcome to design better within Apple's constraints (size, weight, noise) and submit that to them

Adding a heatpipe in the free space we see would not make more noise, make the machine any bigger or weigh significantly more.

So you have the dimensions of the free space available and can suggest a suitably thin heat pipe? There aren't any that will fit in that actual space, there are plenty of high resolution photos for you to work from:
https://www.ifixit.com/Teardown/MacBook+Air+13-Inch+Retina+2018+Teardown/115201
compare the space required to fit such heat pipes in other similar laptops:
https://www.ifixit.com/Teardown/MacBook+Air+13-Inch+Early+2015+Teardown/38266
https://www.ifixit.com/Teardown/MacBook+Air+13-Inch+Mid+2013+Teardown/15042
https://www.ifixit.com/Teardown/MacBook+Air+13-Inch+Mid+2011+Teardown/6130

- I'm sure they'd be happy to hire you.

The constraints they set for themselves and the compromises that result of it are not ones I would agree to work towards.

All products have compromises, they traded off things you don't value as much against something you think is more important. Good for you! Feel free to buy a different product which meets your needs instead of insisting all laptops should be bulky bricks that never thermally throttle. But it seems the market is very happy with thin and light products with reduced compute performance.

[NiHaoMike]

You're welcome to design better within Apple's constraints (size, weight, noise) and submit that to them - I'm sure they'd be happy to hire you.

Easy way to get better cooling is to add thermal pads to use the aluminum case as part of the heatsink. The problem is that the thermal pads, while cheap, do cost something and they try to cut every last fraction of a cent they can get away with.

[wraper]

There isn't space for a heat pipe and associated heatsink in that layout.

Really. If there is enough space for that heatsink, sure there is enough space over PCB for a flat heatpipe. Heatsink adds like 1 centimeter behind the fan, with the same thickness as fan itself.

If you're convinced its so easy, make a cooling package for some of these laptops and sell it to all these people who want better cooling, [sarcasm]it'll be a huge market and you'll be rich[\sarcasm]

One of stupidest arguments. It's not easy to add something to completed design. You cannot simply shrink the fan or cut a hole behind it. Not to say it's a custom fan. But at design stage it wasn't hard to do at all.

[wraper]

You're welcome to design better within Apple's constraints (size, weight, noise) and submit that to them - I'm sure they'd be happy to hire you.

Easy way to get better cooling is to add thermal pads to use the aluminum case as part of the heatsink. The problem is that the thermal pads, while cheap, do cost something and they try to cut every last fraction of a cent they can get away with.

Easy, but then instead of burning itself, CPU will burn your lap because enclosure will become hot.

[langwadt]

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

Adding heatpipe with a heatsink behind the blower fan does nothing to prevent fan from moving air across the PCB from intake side .

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air? 

[BrianHG]

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

Adding heatpipe with a heatsink behind the blower fan does nothing to prevent fan from moving air across the PCB from intake side .

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

The idea here is that the current fan design doesn't even try to move air anywhere in the design, that just a smidge of effort could have improved the situation without added cost, maybe even saving apple money in the long run.

[Someone]

There isn't space for a heat pipe and associated heatsink in that layout.

Really. If there is enough space for that heatsink, sure there is enough space over PCB for a flat heatpipe. Heatsink adds like 1 centimeter behind the fan, with the same thickness as fan itself.

If you're convinced its so easy, make a cooling package for some of these laptops and sell it to all these people who want better cooling, [sarcasm]it'll be a huge market and you'll be rich[\sarcasm]

One of stupidest arguments. It's not easy to add something to completed design. You cannot simply shrink the fan or cut a hole behind it. Not to say it's a custom fan. But at design stage it wasn't hard to do at all.

Your "argument" that with a completely different set of design goals the product could have been designed differently is the silly one. There is no way to fit a heatpipe cooling system in there without adding size and weight. Check the examples linked above of their previous products and see just how much volumetric space is needed, now find that wasted volumetric space in the design. The CPU is one thing that is cooled by the heatsink, but not the only thing.

[Someone]

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

Adding heatpipe with a heatsink behind the blower fan does nothing to prevent fan from moving air across the PCB from intake side .

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

The idea here is that the current fan design doesn't even try to move air anywhere in the design, that just a smidge of effort could have improved the situation without added cost, maybe even saving apple money in the long run.

It looks live a very clearly designed air tunnel to myself and others, air comes in one end, passes across the main board with its heatsink, and is pushed out of the case by the fan. This has been done in all sorts of computers, laptop and desktop, over the years.

[wraper]

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

Because area of that heatsink is very small.

[langwadt]

It might cool the CPU at the other end of the heat pipe better, but what about all the other parts on the PCB? Having the airflow across almost the entire motherboard cools those parts better than the plastic enclosed laptops that carry all the heat around in pipes.

Adding heatpipe with a heatsink behind the blower fan does nothing to prevent fan from moving air across the PCB from intake side .

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

The idea here is that the current fan design doesn't even try to move air anywhere in the design, that just a smidge of effort could have improved the situation without added cost, maybe even saving apple money in the long run.

how do you know it doesn't even try to move air anywhere in the design? if the air intake is on the side opposite the fan the air flows past the heatsink

and a heat pipe would just take up more space to move the heat to a similar sized heatsink exposed to similar air flow just on the the exhaust side of the fan

[langwadt]

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

Because area of that heatsink is very small.

look like a heat sink in the fan exhaust would be of similar size

[Someone]

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

Because area of that heatsink is very small.

look like a heat sink in the fan exhaust would be of similar size

Not even close in volume, the heatsinks on the exits of those fans are tiny in the previous generations. The commenters on here have likely never even handled one of these machines but claim to know the dimensions. Despite the easy references:
https://www.ifixit.com/Guide/MacBook+Air+13-Inch+Early+2015+Fan+Replacement/40700

[SiliconWizard]

You're welcome to design better within Apple's constraints (size, weight, noise) and submit that to them - I'm sure they'd be happy to hire you.

Easy way to get better cooling is to add thermal pads to use the aluminum case as part of the heatsink. The problem is that the thermal pads, while cheap, do cost something and they try to cut every last fraction of a cent they can get away with.

Easy, but then instead of burning itself, CPU will burn your lap because enclosure will become hot.

Yes. It's so easy huh?

[james_s]

I have to think that if the fan were indeed useless, it would have been omitted because fans are not free, they create noise, consume battery life and take up space. Apple has never been known for incorporating gratuitous fans, heck the early Macs had no fan at all even when they really should have.

[NiHaoMike]

Easy, but then instead of burning itself, CPU will burn your lap because enclosure will become hot.

Done right, the bottom of the case will get very warm but not too hot to touch. Or design it to mostly use the top of the case as the heatsink, the part above the keyboard that is rarely touched in actual usage.

I wonder how long before someone comes out with a cost effective thermal transfer material that conducts heat very well at lower temperatures but transition into becoming an insulator above a certain temperature. One way I can think of is a chamber with a carefully controlled amount of alcohol in it, such that it operates as a heatpipe at lower temperatures but go above a certain point and all the liquid vaporizes and ceases to transfer heat well until the temperature drops.

[KaneTW]

In the end, the design choice justification doesn't really matter. What matters is that apparently the CPU keeps burning out on these macbooks, and that's pretty unacceptable for a (iirc) 2 year old model.

[Someone]

In the end, the design choice justification doesn't really matter. What matters is that apparently the CPU keeps burning out on these macbooks, and that's pretty unacceptable for a (iirc) 2 year old model.

What is the rate of incidence? Is it known/proven to be thermally caused? Seems like a massive overreaction and lots of guessing around some insignificant number of failures.

[stevelup]

I honestly cannot see anything wrong with this design at all.

This air tunnel design is perfectly adequate and efficient. To say it's useless is just nonsense. You can see the seals on the bottom of the case that form the tunnel. Air is clearly being drawn in one side of the hinge, across the motherboard / CPU and back out the other side of the hinge. There's even a raised part in the middle to force air right through the fins of the heatsink. It absolutely does not do 'nothing'.

This is just another typical LR clickbait shoutfest!

[wraper]

Easy, but then instead of burning itself, CPU will burn your lap because enclosure will become hot.

Done right, the bottom of the case will get very warm but not too hot to touch. Or design it to mostly use the top of the case as the heatsink, the part above the keyboard that is rarely touched in actual usage.

Regardless how hot it gets, it would suck a lot. I personally would avoid buying a laptop which heats up on the bottom or in keyboard area.

[Black Phoenix]

One thing that Apple could do is exactly what Razer have been doing with his Blade models.

They also for years had thin and light laptops, to compete directly with the Macbook Pro, that thermal throttle like hell, with a design similar to Apple.

Then they started using a Vapour Chamber type of cooling:





The reality now is that the temps are keep at bay, the performance of the laptop is increased because there are more headroom for the CPU to Turbo itself for more time and they pack a RTX2080 Max-Q on it. The clocks from the CPU and GPU goes down yes after some heavy workloads but not as much as similar designs, seeing what it has inside.

https://www.notebookcheck.net/So-much-better-than-before-Razer-Blade-Pro-17-Laptop-Review.424150.0.html

It's only beat by others, more "gamery" laptops, with a sleek look (specially in the Silver option that fits very well in the office environment) and with the 17'' Option that the Macbook killed some generations ago. Problem is that the power adapter is a literally almost a small cinder block...

If Apple wants, they can do better, now do they want?

[thinkfat]

Talking about stupid design, I'm currently repairing a Surface Pro 4 tablet PC. They decided to make the battery part of the cooling concept  .

If you look at the various videos about replacing the battery, you'll see that there's a heat pipe going from the CPU towards the battery. It ends in a large copper plate that is supposed to passively dissipate some of the waste heat, but it is directly layered on top of the battery!

The battery actually has a warning label to not heat it above 140°F (60°C) but who cares, right? 

Guess why I need to repair the tablet? The battery puffed up and created enough mechanical stress to break the adhesive gluing the display into the case...

[tooki]

In the end, the design choice justification doesn't really matter. What matters is that apparently the CPU keeps burning out on these macbooks, and that's pretty unacceptable for a (iirc) 2 year old model.

The subject of this thread is the brand new 2019 MacBook Air, not older versions. (Not that there's any evidence they routinely fail due to overheating, either.)

In the end, the design choice justification doesn't really matter. What matters is that apparently the CPU keeps burning out on these macbooks, and that's pretty unacceptable for a (iirc) 2 year old model.

What is the rate of incidence? Is it known/proven to be thermally caused? Seems like a massive overreaction and lots of guessing around some insignificant number of failures.

Indeed. I don't think there's any evidence that they fail often, never mind due to a single reason. The anecdotes from Rossmann et al (who see only failed units, not the millions that don't fail) can't in any way be extrapolated into a failure rate, as the Apple haters love to do.

why use a heat pipe to move heat to a heatsink in the fan output when you can use a heatsink directly mounted and in the
fan intake air?

The idea here is that the current fan design doesn't even try to move air anywhere in the design, that just a smidge of effort could have improved the situation without added cost, maybe even saving apple money in the long run.

1. The current design clearly does move air around, using the bottom case as an air guide. The airflow doesn't really care where the fan is, what matters is where the flow happens, and it's going over the CPU just fine.

2. You're taking as a given that there is a "situation" that needs to be "improved", and there's exactly zero evidence of this. The TDP of the CPU in this thing is 7W, far less than what was in earlier MacBook Airs (mine has an i7 with 17W TDP), so it doesn't need the same cooling solution.

[GeorgeOfTheJungle]

IIANM, if that MacBook were a solid lump of aluminium, with no fan and 7W its temp would raise by 21 degrees C per hour. IOW 7W is no-thing.

density of aluminium 2.7 g/cm3
specific heat 880 J/kg°K
size 28*20*0.9 cm

(7*3600)/(880*28*20*0.9*2.7/1000)

But I don't believe that 7W figure. It surely eats much more than that in turbo boost and there are other chips there it's not just the CPU.

[GeorgeOfTheJungle]

And the reason people don't like/want to put (much) copper in the laptops is that it's very heavy!

[thinkfat]

And the reason people don't like/want to put (much) copper in the laptops is that it's very heavy!

It's also quite expensive compared to aluminium.

[thinkfat]

It's also quite expensive compared to aluminium.

Apple has made enough money. I've seen PCs using full magnesium construction at similar or even slightly lower price.

Yeah, they have a lot of money in the bank, but they didn't get there by "gilding the lily". Besides, a magnesium alloy construction is a selling point for a laptop, copper heat pipes are not.

[GeorgeOfTheJungle]

But I don't believe that 7W figure. It surely eats much more than that in turbo boost and there are other chips there it's not just the CPU.

This makes my MacBook jump from 22 (1.8 amps/12.3V) to 81 watts (6.6 amps) :

Code: [Select]

nothing added to commit but untracked files present (use "git add" to track)
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[1] 87342
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[2] 87343
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[3] 87344
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[4] 87345
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[5] 87359
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[6] 87360
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[7] 87361
10-YOSEMITE:cosas_iot admin$ cat /dev/zero > /dev/null &
[8] 87362
10-YOSEMITE:cosas_iot admin$ sudo kilall cat
Password:
sudo: kilall: command not found
10-YOSEMITE:cosas_iot admin$ sudo killall cat
[1]   Terminated: 15          cat /dev/zero > /dev/null
[2]   Terminated: 15          cat /dev/zero > /dev/null
[3]   Terminated: 15          cat /dev/zero > /dev/null
[4]   Terminated: 15          cat /dev/zero > /dev/null
[5]   Terminated: 15          cat /dev/zero > /dev/null
[6]   Terminated: 15          cat /dev/zero > /dev/null
[7]-  Terminated: 15          cat /dev/zero > /dev/null
[8]+  Terminated: 15          cat /dev/zero > /dev/null
10-YOSEMITE:cosas_iot admin$

That's what turbo boost does! After a few secs it throttles down back to 5 amps/61W.

Edit:
Got a MacBook Air 8.1, just sitting there doing nothing eats 0.68 amps @ 8.6V or 5.8W, do the cat thing as above and jumps to 1.9 amps/16.3W.

[NiHaoMike]

One other idea is to put the CPU behind the display and use the entire back of the display as a heatsink, with basically a large, flat heatpipe to spread the heat. But unlike a tablet, the battery is still in the keyboard part.

[andy2000]

One other idea is to put the CPU behind the display and use the entire back of the display as a heatsink, with basically a large, flat heatpipe to spread the heat. But unlike a tablet, the battery is still in the keyboard part.

I had an IBM Thinkpad (PIII) that used the display as a heat sink.  There was a heat pipe from the CPU that went to a "thermal hinge" that connected to a large plate behind the LCD panel.  I'm not sure how effective it was, but I remember it being listed as a feature.

[tooki]

But I don't believe that 7W figure. It surely eats much more than that in turbo boost and there are other chips there it's not just the CPU.

TDP is NOT the peak dissipation. It’s absolutely normal for it to dissipate more during peak performance, including turbo boost. (Bear in mind that turbo boost may not even be the highest dissipation; turbo boost is a technique to allow a CPU core to “borrow” thermal capacity from other cores that are idle, allowing that one core to run at a a higher speed than if it had to run alongside other active cores. As such, a multithreaded test like what you did is actually not showing turbo boost, but rather multi-core behavior. All the cores cumulatively may dissipate more than some of the cores in turbo boost.)

Anyhow, the point was that this new model uses a CPU with less than half the TDP of the earlier models it’s being compared to. (Those models’ TDP also wasn’t their peak dissipation, and they also had turbo boost to contend with.) As such, the new model doesn’t need the same cooling infrastructure as the older ones.

[GeorgeOfTheJungle]

I think TDP is the power the cooling system has to be able to dissipate, isn't it? And a MacBook just like that one sitting doing nothing consumes almost 6W already and +16W with the CPU doing something.

[tooki]

I think TDP is the power the cooling system has to be able to dissipate, isn't it? And a MacBook just like that one sitting doing nothing consumes almost 6W already and +16W with the CPU doing something.

Not exactly. The TDP is the maximum expected dissipation under normal use. It can be exceeded at times. And above all, it’s just the dissipation of the CPU, not of the whole system.

https://en.wikipedia.org/wiki/Thermal_design_power states:

The TDP is typically not the largest amount of heat the CPU could ever generate (peak power), such as by running a power virus, but rather the maximum amount of heat that it would generate when running "real applications". This ensures the computer will be able to handle essentially all applications without exceeding its thermal envelope, or requiring a cooling system for the maximum theoretical power (which would cost more but in favor of extra headroom for processing power).

But again, the point is that the new model has less than half the TDP of the older one, so it plain and simply does not need as beefy a cooling solution.

[GeorgeOfTheJungle]

One thing Louis doesn't mention in the video is that the heat pipes' radiator always ends up clogged with debris in the fins, sooner or later. The other design is more resistant in that respect. How many Macs die because of that? Millions I'm sure, because most people don't even know there's a radiator there that needs to be serviced every so often. Not only the laptops, the iMacs too.

[tooki]

One thing Louis doesn't mention in the video is that the heat pipes' radiator always ends up clogged with debris in the fins, sooner or later. The other design is more resistant in that respect. How many Macs die because of that? Millions I'm sure, because most people don't even know there's a radiator there that needs to be serviced every so often. Not only the laptops, the iMacs too.

The reliability stats for Apple computers do not in any way support your absurd claim that millions of them die due to dust-related overheating (or, indeed, due to overheating of any description).

Remember: Louis is in the business of repairing broken devices. For each broken one he sees, he doesn’t see the hundreds that never failed. It’s not a representative sample. And he has a vested interest in making it out as though Apple devices fail left and right, when they don’t.

[GeorgeOfTheJungle]

One thing Louis doesn't mention in the video is that the heat pipes' radiator always ends up clogged with debris in the fins, sooner or later. The other design is more resistant in that respect. How many Macs die because of that? Millions I'm sure, because most people don't even know there's a radiator there that needs to be serviced every so often. Not only the laptops, the iMacs too.

The reliability stats for Apple computers do not in any way support your absurd claim that millions of them die due to dust-related overheating (or, indeed, due to overheating of any description).

Remember: Louis is in the business of repairing broken devices. For each broken one he sees, he doesn’t see the hundreds that never failed. It’s not a representative sample. And he has a vested interest in making it out as though Apple devices fail left and right, when they don’t.

Ohh yeah, because Apples don't even get dirty, right? That only happens to lesser PCs.

[tooki]

False equivalency. I never said they don’t get dirty. But I did say that your hyperbolic claim that they’re failing by the millions because of that is bullshit, because it is.