MacBook Pro 14" vs MacBook Air

[Mike Boreham]

If you look at Geekbench scores, it may seem like the performance drop is only 10%, but I gotta remind you that the M1 Pro with 10 cores, which should be 2x faster in almost every CPU task compared to M1 due to having 2x more performance cores, only scores about 12300 points, which is roughly just 64% faster.

So Cinebench scoring does not scale linearly. In reality, if you're losing only 10% in Cinebench, you may be losing a lot more.

My experience is much closer to this post here.

When the Air throttles due to the GPU-intensive tasks I listed above, the performance drop is closer to 30% or more. Fusion 360 and Room EQ Wizard both become noticeably sluggish, and the machine was very hot to touch.

I am not making any comparison with M1 Pro machines which are obviously in a different league. Nor would I claim that there aren't some usages where M1 Air throttling is more than 10% after 30 mins at max CPU, maybe especially gaming which I don't do and might be a usage where a cooled machine is more appropriate.

I am just picking up on the commonly made dismissive statement that M1 Air's throttle 'quite a bit'.

(BTW I didn't mention Geekbench, which doesn't load the CPU continuously like Cinebench does. Your reply above seems to equate Geekbench and Cinebench).

 

[bill-p]

I am not making any comparison with M1 Pro machines which are obviously in a different league. Nor would I claim that there aren't some usages where M1 Air throttling is more than 10% after 30 mins at max CPU, maybe especially gaming which I don't do and might be a usage where a cooled machine is more appropriate.

I am just picking up on the commonly made dismissive statement that M1 Air's throttle 'quite a bit'.

(BTW I didn't mention Geekbench, which doesn't load the CPU continuously like Cinebench does. Your reply above seems to equate Geekbench and Cinebench).

Ah, sorry, I meant Cinebench. Cinebench scoring isn't linear with CPU cores in M1 family chips for some reason.

And yeah, I think the Air throttles more with GPU-heavy tasks than it does with CPU tasks. I may be wrong, but that's what I'm seeing. And I do realize a lot of my tasks depend on and want a faster GPU. The M1 Pro is kind of an obvious choice for my workflow.

 

[andrewstirling]

I do and wouldn't have it any other way, I see way less detail on low brightness, especially in a lit room.

What you’re doing is massively reducing the dynamic range of the picture and flattening the overall image. You’ll be seeing below black detail you’re probably not meant to see and crushing detail in the brighter parts of the picture.

 

[Love-hate 🍏 relationship]

I'm not getting notebookcheck's results. This is likely due to difference in type of HDR content I'm watching versus theirs. For instance, if there aren't many bright scenes that push whole screen brightness to 1000 nits, I suspect battery life isn't affected as significantly.



The base model gets around 10000 points in Geekbench, and even in Low Power Mode, it's still around 8000.

So it is faster regardless.

There is no flaw. Honestly, you just have to have a MacBook Air to see how slow it gets. In Low Power Mode, the multicore score for the thing gets only around 4500.

Even if Low Power Mode isn't engaged in the MacBook Air, if you push the CPU hard for longer than 5 minutes, it starts throttling by quite a bit due to the lack of any cooling.

Agree on your notebookcheck analysis,very few content gets up to 1600nits

I am sorry but I don't call 8000 faster.m1 is 7777 ...

Yep I have the MBA if u recall (or should I say I had,just sold it lol),too bad I didn't test it in low power mode tho


As for throttling we already discussed about it,and I agree ,the air isn't suited for sustained loads

 

[n3sh]

I'm seeing not so good results for 14" M1pro.
CPU + memory intensive tasks, code build , test runs. I'm seeing M1 Air performing consistently. while 14" even with more core, sometimes takes half the time and sometimes takes 2-3 times as long.

Looks like memory issue in my case for 14", it runs out of memory and then struggles for the rest of the execution.

 

[bill-p]

Agree on your notebookcheck analysis,very few content gets up to 1600nits

I am sorry but I don't call 8000 faster.m1 is 7777 ...

Yep I have the MBA if u recall (or should I say I had,just sold it lol),too bad I didn't test it in low power mode tho


As for throttling we already discussed about it,and I agree ,the air isn't suited for sustained loads

Well, M1 Pro with 8 cores at 8000 is still slightly faster, and that's just CPU score. GPU score still remains a lot higher than M1. I can't be bothered to try and check but honestly, there's not much of a drop in GPU performance going to Low Power Mode. It seems only the CPU is affected.

At this point, we're basically just nitpicking, of course. I'm not saying the Air or 13" Pro are bad machines. Just that they are going to lag behind for certain workloads that the M1 Pro was designed for (high CPU load, high GPU load), and that's just the nature of it all.

In contrast, the Air is fanless and silent, while the 13" Pro lasts almost forever. Also the base MacBook Air is now the best value ever except for gaming and heavy 3D modeling work, but hey, that's what the M1 Pro/Max MacBooks are for. They each have their own benefits. It's good that we have choices.

I'm just saying... with my workflow (and I guess any workflow that's similar to mine), the Air is not enough, and the 13" Pro doesn't provide enough of a benefit in battery life either... to give up on performance.

 

[kingjames1970]

It’s not all about which is the fastest in terms of raw power. The M1 MBA is such a great, light and powerful machine. Brilliant to grab and go in my camera bag - runs Lightroom and PS perfectly. It leaves the house far more than my 15” MBP ever did.

 

[spike1jg]

Thanks for the great information!

 

[bill-p]

I'm seeing not so good results for 14" M1pro.
CPU + memory intensive tasks, code build , test runs. I'm seeing M1 Air performing consistently. while 14" even with more core, sometimes takes half the time and sometimes takes 2-3 times as long.

Looks like memory issue in my case for 14", it runs out of memory and then struggles for the rest of the execution.

This is quite odd. I haven't seen this in my 14" (yet?) but I have seen the discussions around this issue.

I use iStat to monitor CPU Package Power and it seems mine varies between 5 - 16W under light use, or 20 - 35W when stressed a bit.

Memory pressure is very low, though. Under light use, it's about 10%, and even under heavy workload, it's only about 60%.

I don't know if the fact that I have 16GB of RAM has anything to do with it, but I honestly haven't seen any memory leak.

 

[kp98077]

Last night I found my old Air, interesting the size difference, its even larger than the 14"! Am i the only or do others wish the 14" was a tad longer? Oh would be nice to have the apple light up too! LOL

 

[Skewlovevism]

I would also add to that the 50% faster charging, bigger battery and subjectively a better design, as the MBA feels to feminine.

Try to watch THE HU - WOLF TOTEM on YouTube, on your MacBook Air, to increase its testosterone levels

 

[n3sh]

This is quite odd. I haven't seen this in my 14" (yet?) but I have seen the discussions around this issue.

I use iStat to monitor CPU Package Power and it seems mine varies between 5 - 16W under light use, or 20 - 35W when stressed a bit.

Memory pressure is very low, though. Under light use, it's about 10%, and even under heavy workload, it's only about 60%.

I don't know if the fact that I have 16GB of RAM has anything to do with it, but I honestly haven't seen any memory leak.

I've seen it consistently since day 1. First I thought it was spotlight indexing but even now it does that after 2 weeks of usage.

Air executes it in about ~5% of the time consistently. While M1pro struggles depending on the mood.

 

[Queen6]

If you look at Geekbench scores, it may seem like the performance drop is only 10%, but I gotta remind you that the M1 Pro with 10 cores, which should be 2x faster in almost every CPU task compared to M1 due to having 2x more performance cores, only scores about 12300 points, which is roughly just 64% faster.

So Cinebench scoring does not scale linearly. In reality, if you're losing only 10% in Cinebench, you may be losing a lot more.

My experience is much closer to this post here.

When the Air throttles due to the GPU-intensive tasks I listed above, the performance drop is closer to 30% or more. Fusion 360 and Room EQ Wizard both become noticeably sluggish, and the machine was very hot to touch.

Horses for courses; new M1 Pro & M1 Max scale far better with demanding multi core applications. Air & M1 13" still remain to be very fast and extremely efficient notebooks that easily compare with windows based hex core if not octa core notebooks at a fraction of the size, weight and power demand.

Overall the thread seems to have drifted from comparison to a justification of purchase? If I needed a 14" or 16" Pro or Max I'd just buy one as that would make sense, equally the M1 Mac's still remain to be extremely capable and punch well above their weight and present tremendous value for money...

Q-6

 

[throAU]

This is pretty much my experience with low power mode.

Low power mode caps power usage under stress - but does very little if you're lightly using the machine.

My recommendation:
Leave low power mode off unless you're doing heavy work on the machine and want to get a bit more life out of it, or quiet the fan down if it is loud. This will let it ramp performance up as much as required for short-term high stress workloads to maximise responsiveness and performance. But if you're say - doing a render or something long-duration intense and the machine is warmer or louder than you want... set low power mode.

If you're doing light use it makes no difference.

 

[bill-p]

Overall the thread seems to have drifted from comparison to a justification of purchase?

Well, this is normal here though. Many discussions basically come down to "my machine is not as bad as you are making it out to be" or "my machine is better than yours because..."

But yes, I agree. I'm not saying the M1 Air/Pro are bad machines. I'm just saying my particular workflow is a better fit with the 14" M1 Pro (or even M1 Max) and that these are machines made for that kind of use case. But also that I tried a MacBook Air with M1 and it's not able to show me any improvement at all with this workflow.

That's really all it is. I'm pretty sure there are use cases where the Air/Pro will be vastly superior to the 14" and 16".

This is pretty much my experience with low power mode.

Low power mode caps power usage under stress - but does very little if you're lightly using the machine.

My recommendation:
Leave low power mode off unless you're doing heavy work on the machine and want to get a bit more life out of it, or quiet the fan down if it is loud.

If you're doing light use it makes no difference.

Yeah, even on the 14", I think I'd only use Low Power Mode when I am in a very warm tropical place and I want to make sure the device will last a bit longer under load.

The trade-offs: loss of 120Hz mode, lower CPU performance, choppier scrolling, etc... are honestly not worth it to me.

It's just that on the MacBook Air, the mode seems to cause insane performance loss for some reason. It's insane even from the perspective of the MacBook Air itself.

 

[throAU]

As an aside this is exactly what I've always wanted from a machine like this.

The old "discrete GPU on/off" setting sort-of did this on 15" machines, but totally crippled performance because your dGPU was turned off entirely.

Now its excellent. Basically I treat it as a "be quiet" switch if I'm running a game, render or something. Or a "I'm hot" if I have it on my lap doing the same.

But, if I need something completed quicker, letting the fan run will enable that.

 

[Queen6]

Well, this is normal here though. Many discussions basically come down to "my machine is not as bad as you are making it out to be" or "my machine is better than yours because..."

But yes, I agree. I'm not saying the M1 Air/Pro are bad machines. I'm just saying my particular workflow is a better fit with the 14" M1 Pro (or even M1 Max) and that these are machines made for that kind of use case. But also that I tried a MacBook Air with M1 and it's not able to show me any improvement at all with this workflow.

That's really all it is. I'm pretty sure there are use cases where the Air/Pro will be vastly superior to the 14" and 16".

Agree; I would like a 2021 M1 Pro 14" for the features, yet the M1 13" MBP serves my needs better. 2021 16" can, yet not travel friendly due to the footprint.

Q-6

 

[Yebubbleman]

This has been another exciting episode of Apples and Oranges!

 

[Shirasaki]

My workload on M1 MacBook Pro is not heavy but involves 50% of un-optimized applications running consistently in the background for quick access, including a remote access utility that I use to connect to my PC. M1 struggles to handle those applications from time to time. No hiccup but I can feel the machine warming up from time to time. Battery life is nowhere near as advertised or those crazy 18-hr numbers or so, although 6-8 hours of battery life is easily achievable with mixed mid-to-light use and 40% brightness. Active cooling makes all of these possible, cause the M1 MacBook Air runs really warm consistently during my two-week trial of roughly the same workload.

Honestly, I just don't believe Youtuber's MacBook Air "suitable for majority of people" claim, cause M1 is still not efficient enough to push even somewhat CPU/GPU intensive medium workload without getting warm. You can only do so much with passive cooling. Having a fan is highly suggested for people who wants to use MacBook more than just watching youtube videos and scrolling twitter and occasionally doing processing power intensive stuff.

 

[mappleN]

Well, maybe my light usage is much more intensive than I thought. I don't know. Let's just say that's what I'm seeing with the screen on continuously at 60% brightness.

Also, I am not alone:
https://www.reddit.com/r/macbook/comments/lewlc4
And in case you're thinking I'm just making all of this up...

View attachment 1912880

We got the MacBook Air for my sister-in-law. I'm helping her set it up before sending it out.

Just a quick question: Why does you MBA still have the old layout for the arrow keys? Is this a regional thing?

 

[mrchinchilla]

My workload on M1 MacBook Pro is not heavy but involves 50% of un-optimized applications running consistently in the background for quick access, including a remote access utility that I use to connect to my PC. M1 struggles to handle those applications from time to time. No hiccup but I can feel the machine warming up from time to time. Battery life is nowhere near as advertised or those crazy 18-hr numbers or so, although 6-8 hours of battery life is easily achievable with mixed mid-to-light use and 40% brightness. Active cooling makes all of these possible, cause the M1 MacBook Air runs really warm consistently during my two-week trial of roughly the same workload.

Honestly, I just don't believe Youtuber's MacBook Air "suitable for majority of people" claim, cause M1 is still not efficient enough to push even somewhat CPU/GPU intensive medium workload without getting warm. You can only do so much with passive cooling. Having a fan is highly suggested for people who wants to use MacBook more than just watching youtube videos and scrolling twitter and occasionally doing processing power intensive stuff.

It will differ from case to case. My M1 Air has been perfect for me and my work, which isn’t overly CPU intensive at all - mostly proprietary Windows software running under Parallels, and Word, Excel, etc.

I haven’t had any issues with heat, temps rarely ever go above 40 Celsius, and it never gets even slightly warm to the touch during typical use.

I get an average of 8 hours battery life with that workload at 50% brightness, and it can be pushed past 10 hours if low power mode is turned on, which oddly doesn’t have that big of an effect on performance for me unlike what I’ve seen some saying on here, even when running more intensive tasks like database building (albeit in emulated x86 software running on Parallels), I’m sure it’s more noticeable doing more CPU expensive tasks, but it’s still magnitudes faster than it was running natively or virtualised on my old 2013 MBP.

If I were to buy a laptop now, I’d probably spend the extra and go for a 14” M1 Pro, although I don’t like the sharp edges on the palm rests (like the old MBP design, it really digs into your palms of you do a lot of typing, and eventually gets pattina’ed and worn away), the tapered design of the MBA is far more comfortable. But the MBA was the best option at the time when I needed a new computer, and the new MBPs were still nothing but conjecture.

For anyone who doesn’t need/doesn’t want to pay the extra for the performance of the new MBPs, the M1 Air is definitely a very good option, but if I were in that position now I’d probably just wait for the new M2 Air likely to be released in the first half of next year.

 

[Thisismattwade]

I appreciate the OP, and what he was trying to say. We have the base M1 MBA. It has been a wonderful machine, and we even have it setup with two users: my wife and myself. The screen and keyboard of the 14" MBP are enticing to me, but we literally have no use for most of that power/performance. We charge the MBA maybe once or twice a week, and neither of us are "pro" users in any sense of the word. (I don't even know what Light Room looks like!)

For a comparison: for our workflows it goes toe-to-toe with my brand new, work-issued Dell XPS 13" (i7-1185, 16GB RAM). Crunching numbers in Excel, several Edge browser tabs open with different sites (& Pandora playing music), MS Teams (calls and meetings sometimes), a Citrix app or two occasionally. All of this with the icing on the cake: I snagged the MBA for around half the Dell: $760 vs $1350.

I still recommend the (refurbished) base MBA all the time, as I don't think the next-gen MBA is going to be as cheap as the current version.

 

[theorist9]

One key benefit of the M1 Pro MBP I'd add to your list is that it can drive two external monitors, while the M1 Air can drive only one.

The M2 Air may be able to drive two externals--we shall see.

 

[bill-p]

Just a quick question: Why does you MBA still have the old layout for the arrow keys? Is this a regional thing?

I think you're thinking of the MacBook Air with butterfly keyboard (up to 2018-2019). Ever since Apple ditched butterfly keys, the arrow keys have always been this small in the US layout. I have seen some other layouts and they also seem to have the arrow keys at this size.

 

[ADGrant]

I get that's what you're seeing and I don't mean to imply you're faking it. My M1 MBA has had 18-20 hour battery life for my workflow which is web work, VNC client to VMs and such. I don't generally allow this machine to have any type of CPU load as I know that is a killer to battery life. I use macOS Stats app to track so keep an eye out to see if anything is causing your machine to have high CPU load.

If you are using your MacBook as a chromebook/thin client then yes the base MBA should be more than sufficient. Your 18-20 hour battery life might not be a reasonable expectation for others.