2020 Air Heatsink Modification Thread

[Mopar]

what do you think a directly mounted air cooler is made off, heat pipe(s)
i'm not suggesting fitting a noctua d15 in there, just take a look at a 2015 mba that has been used for reference by previous users.

by using a heat pipe like the 2015 mba, you transfer the heat directly infront of the fan exhaust to get direct static pressure from the fan. therefore the fan would be much more effective at wicking away the heat than the heatsink being placed couple inches away from the fan's blowing direction.

you guys are really making this way more complicated than it is.

What about the dynamic pressure and temperature gradient? Static pressure is inverse to dynamic pressure (Bernoulli's Principle) and it is dynamic pressure (airflow) that facilitates heat exchange. Static pressure is what the fan works against (molecular resistance). An increase in static pressure = an increase in temperature. An increase in air temperature reduces the temperature gradient, meaning heat transfer is less efficient.

You are making an assertion – that one system is more efficient than another – based on no evidence at all. You have taken none of the variables of surface area transfer, molecular movement and temperature gradient into account. You cannot categorically state that a push fan is more efficient than a pull fan until you have measured the number of air molecules passing over a conductive surface area in a given amount of time. Dynamic pressure and air density (temperature and atmospheric pressure) are what determine the number of air molecules flowing over a surface in any given time.

With respect, I would like to read your understanding of the principles of heat exchange before I debate your claims any further. Because what you have written proves you don't appear to fully understand all the variables.

 

[Mopar]

Seriously, what do you not understand.

Your point.

That is to say, how did you arrive at the arbitrary figure of 3.5GHz sustained clock speed? If the CPU were designed to operate at a sustained 3.5GHz, then it would be a 3.5GHz CPU with (or without) a rated boost speed in excess of this.

But it is not. It is a 1.1GHz (rated) CPU that is able to triple its clock speed for short bursts and, by the looks of it, almost double its clock speed for a sustained period.

Except for my single fan heat piped MBP cools my CPU pulling 20W+. This MBA has to throttle down to pulling 10W. That's the problem. As you say, wattage goes to heat. And tell me why my MBP can dissipate 10W more heat? Heat pipe. End of story.

I have not seen the tests you have conducted, so I can only take them at face value. I do know that the MBA and MBP are not excactly the same form factor. One is heavier than the other and also has more case surface area, which means it will transfer and dissipate heat slightly more efficiently.

Energy can be converted to heat or work. The efficiency of a CPU is based on how much energy is converted to work instead of heat. The caveat is, that work is eventually turned into heat. However, as work transfers energy from one area to another, it also transfers the potential heat from one area to another.

What this means is a CPU may consume X Watts, but only a proportion of X is converted to heat at the CPU level – and that is what the heatsink is ducting away.

So at face value, 20W is 20W – that 20W of heat will be dissipated somewhere at some time. But it does not mean all of it is dissipated via the heatsink on the CPU.

Also, did you calculate the amount of work each CPU was doing whilst they were both running at their various sustained clock speeds? Because if you didn't, then again it is not a fair comparison.

In any case, without measurements we are both speculating which system is more or less efficient and by how much. For that reason, I'm not stating one way or the other which is the more efficient system. What I am stating is that the current MBA system is not what many people on here are claiming it is: a passive cooling system made up of a fan with no direct thermal path from the heatsink.

 

[throAU]

But it is not. It is a 1.1GHz (rated) CPU that is able to triple its clock speed for short bursts and, by the looks of it, almost double its clock speed for a sustained period.

In addition to this: intel has changed the way boost works in the past few years. Both in terms of how they define it in terms of power draw, and the various ways it can boost individual cores.

Comparing "boost" between processors made in 2016 vs. processors made in 2019-2020 on 10nm is chalk and cheese. You are literally comparing two different features.

Even considering the different definition of how boost works.... the manufacturing process is different (and its well known that even given equal cooling, intel 10nm just simply doesn't clock as high as 14nm - but they can fit more cache/cores and iGPU EUs in same/less space - to get performance that way), the design is new.

What matters is how the end product you can buy performs, and the 2020 MBA > previous model (i'd also wager it is competitive or faster at some things than the 2019 13" Pro - because of the new architecture and more powerful integrated GPU). High clocks are not really desirable if you can get the same or better performance at lower clock. Because energy draw goes up exponentially with clock rate. So - comparing clocks and whining that one clocks lower like it is the only metric that matters - is brain damaged.

What Apple could or could not have done is irrelevant. The product people are wishing for (whether legitimately, or out of misguided fantasy) does not exist, and will not exist.

Get over it

 

[magbarn]

I’m surprised none of the youtubers have done the MacBook in the freezer test yet. Cooling it down to couple of degrees below zero will tell us if it’s power constrained or truly heat constrained.

 

[PBG4 Dude]

I’m surprised none of the youtubers have done the MacBook in the freezer test yet. Cooling it down to couple of degrees below zero will tell us if it’s power constrained or truly heat constrained.

I was wondering about this too, since everyone was doing it for the first 6-core MBP models.

 

[adrianlondon]

This is great. Just when I start to get used to the performance of the MBA, Winter will come along and it'll speed up!

 

[raymanh]

With respect, I would like to read your understanding of the principles of heat exchange before I debate your claims any further. Because what you have written proves you don't appear to fully understand all the variables.

With all due respect, you're placing the burden of proof on someone else, despite your argument coming from 'analysis' of 3D airflow from looking at a few pictures, citing some basic thermodynamic principles, and then trying to figure out if a heat pipe would be better 😂.

Let's stick to empirical results shall we? See this below.

That is to say, how did you arrive at the arbitrary figure of 3.5GHz sustained clock speed? If the CPU were designed to operate at a sustained 3.5GHz, then it would be a 3.5GHz CPU with (or without) a rated boost speed in excess of this.

Because the i5 in the new Air goes up to 3.5GHz. I think you need to learn what 'arbitrary' means. Again, the CPU in my 2016 MBP runs at 2.9 Ghz for hours while I game. It stays cool, so clearly it was designed to run at that speed, but it's still rated as a 2.0 GHz base CPU, not a 2.9 GHz CPU as you are suggesting it should be.

I have not seen the tests you have conducted, so I can only take them at face value.

See this screenshot of my 2016 MBP nTB. Specs again; 2.0GHz to 3.1 boost, dual core, single fan heat pipe cooling. It was taken near the end of a Cinebench run so everything is in equilibrium. 19W, 2.89 GHz, 92 deg C and fan speed at 4500 rpm. Look at the graphs, its rock steady at those numbers, no thermal throttling. Now compare that to the 2020 i5 MBA in this video. I actually misread it before; it didn't throttle down to 2 GHz, but actually 1.4 GHz.

I do know that the MBA and MBP are not excactly the same form factor. One is heavier than the other and also has more case surface area, which means it will transfer and dissipate heat slightly more efficiently.

Wrong. How many times do I need to say they are the same size! Look on the Apple website if you still don't believe me. The MBP and 2020 MBA have the same footprint of 30.41 x 21.24 cm. The MBP has a uniform thickness of 1.49 cm while the MBA's is 0.41 – 1.61 cm. The MBA is 1.29 Kg while the MBP is 80 g heavier. The MBP has no cooling advantage here.

What I am stating is that the current MBA system is not what many people on here are claiming it is: a passive cooling system made up of a fan with no direct thermal path from the heatsink.

That's not what we're saying. We're saying it's thermally compromised, that it could have very easily been better if Apple hadn't purposely handicapped it so it didn't overlap with the MBP.

What matters is how the end product you can buy performs, and the 2020 MBA > previous model (i'd also wager it is competitive or faster at some things than the 2019 13" Pro - because of the new architecture and more powerful integrated GPU). High clocks are not really desirable if you can get the same or better performance at lower clock. Because energy draw goes up exponentially with clock rate. So - comparing clocks and whining that one clocks lower like it is the only metric that matters - is brain damaged.

Again let's stick to empirical evidence rather thinking you know, speculating from, no offence, some basic understanding of computers.

Faster than a 2019 13" Pro? Let's see if the MBA can beat my 2016 MBP first.

In the blue corner:
My completely base spec 2016 MBP nTB. Single fan, heat piped, 2.0 GHz 6th gen dual core i5

In the red corner
The new 2020 1.1 GHz 10th gen quad core i5 with G7 graphics.

Round 1. Cinebench R20. (CPU test)

2020 MBA; 863 (proof) vs 2016 MBP; 774 (proof)
MBA 11% higher score.

Round 2. Unigine Heaven 4.0, extreme preset. (Graphics test)

2020 MBA; 209, average FPS 8.3 (proof) vs 2016 MBP; 215, average FPS 8.5 (proof)
MBA is 3% SLOWER!!! It has G7 graphics and it's slower.

So yeah, competitive with the 2019 MBP? I don't think so.

The point here is the MBA could be much faster if it was cooled properly. What is the point of having G7 graphics and a quad core in it?

What Apple could or could not have done is irrelevant. The product people are wishing for (whether legitimately, or out of misguided fantasy) does not exist, and will not exist.

Could've easily existed, heat pipe was omitted to save the MBP. It's lazy from Apple, that's the point.

Get over it

Says the one who's arguing too 🤣

 

[Tubamajuba]

Round 2. Unigine Heaven 4.0, extreme preset. (Graphics test)

2020 MBA; 209, average FPS 8.3 (proof) vs 2016 MBP; 215, average FPS 8.5 (proof)
MBA is 3% SLOWER!!! It has G7 graphics and it's slower.

So yeah, competitive with the 2019 MBP? I don't think so.

The point here is the MBA could be much faster if it was cooled properly. What is the point of having G7 graphics and a quad core in it?

I do want to point out that the i5-6360U in the base 2016 MBP has 64 MB of eDRAM for the GPU to use, whereas the
i5-1030NG7 in the 2020 MBP has no eDRAM. Furthermore, but less important, the 6360U is designed to run at 15W while the i5-1030NG7 is designed to run at 10W.

 

[raymanh]

I do want to point out that the i5-6360U in the base 2016 MBP has 64 MB of eDRAM for the GPU to use, whereas the
i5-1030NG7 in the 2020 MBP has no eDRAM. Furthermore, but less important, the 6360U is designed to run at 15W while the i5-1030NG7 is designed to run at 10W.

Didn't know the first point. Regardless the point of the comparison was to refute throAU's statement that the 2020 MBA could be competitive against 2019 MBPs.

 

[KoolSafak]

MBP 2019 > MPA 2020

I'm disappointed

 

[mick2]

Let's stick to empirical results shall we?


Again let's stick to empirical evidence rather thinking you know, speculating from, no offence, some basic understanding of computers.

Faster than a 2019 13" Pro? Let's see if the MBA can beat my 2016 MBP first.



Round 1. Cinebench R20. (CPU test)

2020 MBA; 863 (proof) vs 2016 MBP; 774 (proof)
MBA 11% higher score.

Round 2. Unigine Heaven 4.0, extreme preset. (Graphics test)

2020 MBA; 209, average FPS 8.3 (proof) vs 2016 MBP; 215, average FPS 8.5 (proof)
MBA is 3% SLOWER!!! It has G7 graphics and it's slower.

So yeah, competitive with the 2019 MBP? I don't think so.

Didn't know the first point. Regardless the point of the comparison was to refute throAU's statement that the 2020 MBA could be competitive against 2019 MBPs.

Some further empirical figures from the YouTube guy you're quoting above:

2020 Air
Geekbench5 1183/3231
Geekbench5 Metal 9390

2019 MBP
Geekbench5 938/3960
Geekbench5 Metal 3165

...so in fact it looks like throAU was correct; the 2020 Air is actually competitive or faster in some areas with the 2019 MBP.

My takeaway is that the Air actually is competitive with (and in some areas) even outperforms a recent MBP in my most common scenarios (single core burst workloads, desktop graphics, IO read & write speeds), at a lower cost and weight, and with likely longer battery runtime. It is thermally more limited (slower) than recent MBPs when running extended, continuous multi-core workloads; even then its multi core score is pretty darn respectable given that it's a low power, thermally limited 10w chip. So yer pays yer money and makes yer choice.

 

[JTToft]

MBP 2019 > MPA 2020

I'm disappointed

- I think that video showed exactly what you could expect.

 

[Pugly]

The 2020 MBA performance compared to 13" Pro is basically what the performance difference has always been like. The 2018/2019 performance was pretty awful in comparison... you gave up a lot of performance for not much else... you didn't even save that much money.

This Air is pretty fast now, as long as you don't need it to export/render/compile something quickly.

I'm curious how good the i7 is. I felt that upgrade got the Air pretty close to a base Pro in the non-retina Airs.

 

[Mopar]

With all due respect, you're placing the burden of proof on someone else, despite your argument coming from 'analysis' of 3D airflow from looking at a few pictures, citing some basic thermodynamic principles, and then trying to figure out if a heat pipe would be better 😂.

Let's stick to empirical results shall we? See this below.



Because the i5 in the new Air goes up to 3.5GHz. I think you need to learn what 'arbitrary' means. Again, the CPU in my 2016 MBP runs at 2.9 Ghz for hours while I game. It stays cool, so clearly it was designed to run at that speed, but it's still rated as a 2.0 GHz base CPU, not a 2.9 GHz CPU as you are suggesting it should be.



See this screenshot of my 2016 MBP nTB. Specs again; 2.0GHz to 3.1 boost, dual core, single fan heat pipe cooling. It was taken near the end of a Cinebench run so everything is in equilibrium. 19W, 2.89 GHz, 92 deg C and fan speed at 4500 rpm. Look at the graphs, its rock steady at those numbers, no thermal throttling. Now compare that to the 2020 i5 MBA in this video. I actually misread it before; it didn't throttle down to 2 GHz, but actually 1.4 GHz.



Wrong. How many times do I need to say they are the same size! Look on the Apple website if you still don't believe me. The MBP and 2020 MBA have the same footprint of 30.41 x 21.24 cm. The MBP has a uniform thickness of 1.49 cm while the MBA's is 0.41 – 1.61 cm. The MBA is 1.29 Kg while the MBP is 80 g heavier. The MBP has no cooling advantage here.



That's not what we're saying. We're saying it's thermally compromised, that it could have very easily been better if Apple hadn't purposely handicapped it so it didn't overlap with the MBP.



Again let's stick to empirical evidence rather thinking you know, speculating from, no offence, some basic understanding of computers.

Faster than a 2019 13" Pro? Let's see if the MBA can beat my 2016 MBP first.

In the blue corner:
My completely base spec 2016 MBP nTB. Single fan, heat piped, 2.0 GHz 6th gen dual core i5

In the red corner
The new 2020 1.1 GHz 10th gen quad core i5 with G7 graphics.

Round 1. Cinebench R20. (CPU test)

2020 MBA; 863 (proof) vs 2016 MBP; 774 (proof)
MBA 11% higher score.

Round 2. Unigine Heaven 4.0, extreme preset. (Graphics test)

2020 MBA; 209, average FPS 8.3 (proof) vs 2016 MBP; 215, average FPS 8.5 (proof)
MBA is 3% SLOWER!!! It has G7 graphics and it's slower.

So yeah, competitive with the 2019 MBP? I don't think so.

The point here is the MBA could be much faster if it was cooled properly. What is the point of having G7 graphics and a quad core in it?



Could've easily existed, heat pipe was omitted to save the MBP. It's lazy from Apple, that's the point.



Says the one who's arguing too 🤣

There is so much misinformation and dishonesty in this post I don't know where to start.

So I will refrain from posting a bunch of boring dot points and simply highlight the most obvious flaw on which all your assumptions are based: this video

If that is the best you have got, this debate is over. No-one takes that video seriously and one day you will figure out why. Suffice it to say, the figures you posted above are based on a new MBA out of the box being stress-tested while it is still indexing and running background programs as part of its initial OS installation.

Have you seen his latest video? Clearly not. The "new" Cinebench 2.0 score is now 955. And he was clearly too embarrassed to run Unigine Heaven 4.0 again after realising his rookie error in the first video.

In addition to this, you continue to argue that your 2-core 2016 MPB is somehow superior to the 4-core 2020 MBA when even a precursory glance shows it's Geekbench 5 score of 1609 is less than half the 3200+ the latest MBA has been scoring.

Please.

You can continue arguing for a heatpipe based on your own flawed analysis – just like you did in your very first post here – but I don't think many reading this are going take your arguments seriously now that they know they are all based on an amateurish YouTube video by someone who clearly doesn't know how to benchmark.

 

[raymanh]

There is so much misinformation and dishonesty in this post I don't know where to start.

Says you who has claimed multiple times the MBA and MBP have different footprints and that's why the MBP inherently cools better.

Says you who has tried to argue points based on a few photos and some basic thermodynamics. Laughable.

So I will refrain from posting a bunch of boring dot points and simply highlight the most obvious flaw on which all your assumptions are based: this video

Fair point, his newer numbers are much higher. I didn't think background tasks while the OS was setting up would have had such a large impact.

Have you seen his latest video? Clearly not. The "new" Cinebench 2.0 score is now 955. And he was clearly too embarrassed to run Unigine Heaven 4.0 again after realising his rookie error in the first video.

955? From a quad core i5. My dual core got 774. Why's it not higher? Cooling.

In addition to this, you continue to argue that your 2-core 2016 MPB is somehow superior to the 4-core 2020 MBA when even a precursory glance shows it's Geekbench 5 score of 1609 is less than half the 3200+ the latest MBA has been scoring.

Please.

In some ways yes, my point is that the cooling in my 2016 MBP is adequate for the CPU. In the 2020 Airs, it's not. Hence why my old MBP can perform fairly close to a new 10th gen quad core in the Air when thermals come into play.

I'll come back to the point on the Geekbench scores at the end of this post, because that's important.

You can continue arguing for a heatpipe based on your own flawed analysis – just like you did in your very first post here – but I don't think many reading this are going take your arguments seriously now that they know they are all based on an amateurish YouTube video by someone who clearly doesn't know how to benchmark.

Right, but you think people take your 'analysis' seriously? Based off you looking at a few pictures, quoting some first year thermodynamic equations thinking and then asking others to justify how a heat pipe would be better?

Yes as I said above it's true the benchmarks are a lot better now for the 2020 MBA. But that was just a side point to what throAU was saying about the 2020 MBA > 2019 MBP.

You know the real argument here is that the cooling in the 2020 MBA is inadequate.

In his new video you can clearly see how quickly the temps in the MBA go up to 100 deg C while the MBP stays cool. That is all you need to know. Look at this video and this video too. In all their tests you see the MBA go to 100 deg C, start throttling, reduce clock speed and only pull slightly above their rated TDP. My MBP doesn't do this.

Additionally, and here's a major point, you showed that in Geekbench 5 the new MBA is 2x the performance (3200+ vs 1609) of my 2016 MBP, while in Cinebench R20 it's only 1.3x the performance (955 vs 774). Why? because Cinebench puts a continuous load on the CPU hence tests its thermals while Geekbench 5 only intermittently loads the CPU so thermals are never tested.

Don't believe me again? On my MBP... Cinebench R20 vs Geekbench 5. Look at the graphs. In Cinebench the load is continuously high, the fans come at the start and it tests the thermals throughout the test. Now look at Geekbench 5. Intermittent loading, the temps gradually climb to 80 deg C while the fan is off, then the fan comes about 10 seconds before then end of the test. Geekbench 5 is all about short bursts. Cinebench tests thermals. The difference in scores literally shows you how bad the heatsink cooling design is in the Air.

You still seriously think the MBA is not flawed thermally? Omitting the heat pipe (which does cool better) was obviously done to differentiate it with the MBP. Can't believe you don't realise that.

 

[Dulcimer]

Can we get a heat pipe for this thread?
👈😎👈

I’m here all night, folks.

 

[PianoPro]

Can we get a heat pipe for this thread?
👈😎👈

I'm rather entertained watching 2 people argue about how to "fix" something that works as Apple intended.

 

[russell_314]

It comes down to if you want better performance buy a MacBook Pro. If you want the smaller form factor buy a MacBook Air. Apple isn’t going to make the MacBook Air into a MacBook Pro regardless of how many people complain on MacRumors forms.
[automerge]1585124166[/automerge]

I'm rather entertained watching 2 people argue about how to "fix" something that works as Apple intended.

Very true. I’m so glad Apple doesn’t listen to people complain on forms as advice on what to do. They have people with college degrees designing these products

 

[jwalesh96]

The way I see it.
The thermal design for pure sustained performance is just bad, the comparisons and videos show it as such and the numbers dont lie. Thats undeniable at this point. However it was most likely to have been designed like this to differentiate from the MBPs and I guess apple is aiming it at people who do spike workloads and want something quiet. Since my workload is a mix of both I prefer a mix of benchmarks for comparisons. Thus I only think of geekbench as another reference point for spike loads and nothing more. Wouldn't use it by itself to draw any conclusions. That and I wouldnt bother with these and would go for a MBP or any other laptop with better cooling. So I understand this isnt aimed at me.

However the disappointment that most people have is that it could be so much more as the previous MBA design had better cooling and now that they've added a quad core option with a bunch of advertisements of performance improvements, people were hoping this would be a perfect machine or if not, are disappointed when it could be so much more. That and the quad core doesnt seem have the headroom to reach its true potential at all.

But will it be fine for most people with light work loads and deal with quick spike loads? probably. Could it be better designed ? Definitely. There is no need to defend em. Apple makes mistakes as well, rather we should all hope they come up with better designs that can also maximize what we get for our money.
Will these Sell? Most likely.

Another Day, Another refresh.
I say, Buy and vote for what you want with your money.

Oh and btw, I'm in the camp where having a degree in something doesn't automatically = Correct. You meet enough people and you'll find varying degree of ability regardless of education. Pulling theory (without testing)and paper qualifications into an argument also doesnt mean anything imo.

 

[raymanh]

The way I see it.
The thermal design for pure sustained performance is just bad, the comparisons and videos show it as such and the numbers dont lie. Thats undeniable at this point. However it was most likely to have been designed like this to differentiate from the MBPs and I guess apple is aiming it at people who do spike workloads and want something quiet.

[...]

However the disappointment that most people have is that it could be so much more as the previous MBA design had better cooling and now that they've added a quad core option with a bunch of advertisements of performance improvements, people were hoping this would be a perfect machine or if not, are disappointed when it could be so much more. That and the quad core doesnt seem have the headroom to reach its true potential at all.

Yep, you've exactly summed up my feelings.
[automerge]1585126035[/automerge]

I'm rather entertained watching 2 people argue about how to "fix" something that works as Apple intended.

I think both sides of the argument agree it works as Apple intended. I mean, I'm saying that Apple intended to not make the new Air be competitive with the pro by making the Air thermally compromised.

The argument is:

Disappointed that the i5 MBA could've been so much better with a proper cooling, i.e. with a heatpipe like the Pro (Me and others)

versus.

A heat pipe wouldn't improve it and the thermals of the MBA are satisfactory (Mopar and others)

 

[raymanh]

Oh and btw, I'm in the camp where having a degree in something doesn't automatically = Correct. You meet enough people and you'll find varying degree of ability regardless of education. Pulling theory (without testing)and paper qualifications into an argument also doesnt mean anything imo.

😂 tell that to Mopar!

 

[PianoPro]

It comes down to if you want better performance buy a MacBook Pro. If you want the smaller form factor buy a MacBook Air. Apple isn’t going to make the MacBook Air into a MacBook Pro regardless of how many people complain on MacRumors forms.

Precisely. The day they want to make a $1000 MBA perform like a $1300 MBP is ... never. People that think the MBA thermal design is a mistake apparently have no idea how market differentiation works to maximize profit. Just because Apple could, doesn't mean Apple should.

The MBA and MPB address competition in different markets. The MBA is sold as the lowest price (portable) to get MacOS, limited to tasks that don't require sustained performance. The MBP is sold as MacOS for more demanding tasks that require sustained performance. Apple is never going to lower their profits by selling a $1000 MBA with the performance of a current $1300 MBP.

 

[russell_314]

Precisely. The day they want to make a $1000 MBA perform like a $1300 MBP is ... never. People that think the MBA thermal design is a mistake apparently have no idea how market differentiation works to maximize profit. Just because Apple could, doesn't mean Apple should.

The MBA and MPB address competition in different markets. The MBA is sold as the lowest price (portable) to get MacOS, limited to tasks that don't require sustained performance. The MBP is sold as MacOS for more demanding tasks that require sustained performance. Apple is never going to lower their profits by selling a $1000 MBA with the performance of a current $1300 MBP.

Well it's about profits of course but also the Air is thinner and lighter. If I need a portable computer for checking my emails, editing office documents, posting comments on forums and other light uses the Air will do it just fine and weigh less in my briefcase. Of course Apple could likely do some modifications to the Air to increase performance without increasing weight but that would also increase the price. Why would I spend extra $$$$ on an Air Pro(?) when I don't need that much power. Those that want Pro power will just buy the MacBook Pro

 

[imp3rator]

Air is better or cheaper than last year model and not as good as base Pro model. It is only about product line-up. Of course Apple could make few things better but it will cannibalize own higher models. That's all...

 

[PianoPro]

Those that want Pro power will just buy the MacBook Pro

If that were true no one would be bitching about the MBA performance all over this thread and a bunch more. They blame Apple for a poor thermal design because they don't want to buy a MacBook Pro. They won't, or at least they think they shouldn't have to buy a MacBook Pro to get Apple's defined "MacBook Pro performance".

But it is NOT a design flaw, it's obviously by design. Simply because they could cool it better (this isn't rocket science) doesn't mean they should. It's product differentiation to maximize their profits. They want people that need sustained performance to buy the more expensive MacBook Pro. They are not going to "fix" the MBA to make it perform like a MacBook Pro because it isn't broken, it's performing at a level they want.

The reason this upsets people so much is because they know the hardware inside the MBA is capable of more performance with a different cooling design that they assume wouldn't substantially raise its manufacturing cost. So they reason it could have more performance for very little if any added price. Hence they complain it's poorly designed. But it isn't Apple's intention to maximize the performance of its lowest priced product. They want you to buy the more expensive MacBook Pro to get more performance which increases their profit.