Apple NEVER Learns. - M2 Macbook Review

Apple’s M2 SoC has been out for a while, but are the new MacBook Air and Pro models living up to the hype? Or is it just a lot of hot Air?

Buy a MacBook Air M2:

Purchases made through some store links may provide some compensation to Linus Media Group.

FOLLOW US
---------------------------------------------------

MUSIC CREDIT
---------------------------------------------------
Intro: Laszlo - Supernova

Outro: Approaching Nirvana - Sugar High

CHAPTERS
---------------------------------------------------
0:00 Intro
1:09 Apple's Cooling Solution (or lack thereof)
2:52 Thermal Testing
3:22 Design, IO and Charging
4:24 Storage
5:50 Benchmarking
7:27 Real World Workloads
8:25 Overall Performance Results
9:17 Battery Life
10:13 Concluding Thoughts on the Air
10:38 As for the Pro...
11:38 Final Thoughts

[dgsprysoup]

Linus, a good idea next time would be to include code compile times for comparisons! Macs are used by many devs and a difference between the normal and Pro models is going to be huge

[ashuggtube]

“It’s always kind of funny to me when people complain about gaming performance on a machine that was clearly engineered for browsing Facebook” 😂

[fandacy]

I have been using base model m2 air for a year now. For the ones that are wondering, It never got hot (not even close to warm) even once during my workflow. Just to be fair my workflow is relatively light. I run these programs at the same time while working: discord, safari with 10ish tabs, onenote, mail, calendar, notes, asana. Hope this helps.

[DanBires]

Don’t forget if that dell xps isn’t plugged in it won’t compete with the macs. The macs score the same on just battery power. The xps scores way lower unplugged.

[JeffGeerling]

I've been using an M2 Air for a month or two now and honestly, I was slightly happier with the M1 Air, for some reason it felt smaller, maybe it was the curves... but it was so good in that form factor that the slight performance upgrade in the M2 coupled with sometimes throttling behavior when I'm video editing made the M2 Air a little disappointing.

But that's just like, my opinion man.

[killertoast]

Would have been cool to include a Ryzen 6000 Ultrabook. These chips are much more power efficient than intel and in a small chassis with limited tdp this makes a big difference in performance

[ohnoimissed]

An M1 MBA was my first ever Apple laptop, and I loved it. Traded it in for an M2 MBA, and honesty, I've hit zero issues so far. The battery life remains fantastic, it rips through the workloads I throw at it (web dev, music, browsing, remote desktop for work, very light gaming, Photoshop/light creativity work).

I had read about the heat issues, so I've been running temperature monitoring software. Vast majority of the time, the SOC is running between 35-40c. There are occasional, short lived spikes in the double digits if the CPU/GPU gets really stressed, but the only time I've seen sustained high temperatures - by which I mean lasting more than a few seconds - are when running games, the most dramatic of which was a brief spike up to 108c before it throttled back and settled into high double digits.

My experience (your mileage may vary, ofc) is that it is probably, for most people, going to be pretty much a non-issue. This thing ain't running hot all the time unless you're using it for stuff that you really should be doing on a pro instead.

[grittygirlgraphics]

I use Photoshop, Illustrator and After Effects and this little M2 Air handles all of it amazingly well. When I migrated from my big machine, the air got really hot, and I was worried. But since then, it hasn't skipped a beat. I love this little portable mini version of the big boy I keep at home.

[MrDeaf]

I remember when macbooks had active cooling. They would break their hinges because all the heat was exhausted into the structural glue on the screen.

[KingMuttley]

The best thing about Apple having shite cooling is that we can get to see Alex go all mad scientist on it

[michaelhonaker]

You should try container benchmarks. Alot of developers use Mac with docker and I'd like to see it's performance is. I know it's arm so won't be comparable to any competing x86 platform

[oreo_man]

I there any chance you guys could build your own cooling system for one of these to see how well the M2 could perform if it had adequate cooling?

[Vikushu]

The main similarity between every Macbook Air M2 review I've seen is they never recommend getting the base spec. But if you ask me, if you were going to get the base M2 Air, you might as well get the base M1 Air since it gets most of the way there for common use cases, and also isn't crippled in read/write performance.

[northgen]

It would be great to compare « real world performance » with other tasks. Not everyone is a film maker, some of us are engineers, accountants, programmers, etc. So knowing if a laptop can run simulations, conception work, big macros and excel sheets, some heavy code would be very interesting!

[mattifine]

It would be interesting to se what the Mac air could do if not thermally limited. With some DIY water cooling.

[cputeq]

Bought a 1TB, 16GB M1 Air (8/8) refurb from Apple ($300 less than retail and Apple refurb has always been brand-spanking new for me). Great little machine -- I love the integration between my iPhone and MacOS.

Interesting side note (talk about benchmarking) - I performed the M1 Air thermal pad mod (which allows the chassis to REALLY get hot) and saw definite performance improvements in sustained workloads.

Then, I got curious to see just how hot the CPU would get. I had a thermal monitoring program running (which even shows battery temps) while playing Minecraft (with shader modes) for 30 minutes or so. The machine got extremely toasty (I had it on my desk) as the chassis soaked up all the heat.

Then, I did some digging on the 'net to see what the 'safe' temperature is for Li-Ion batteries, and found an interesting tidbit -- they have a higher temp headroom during discharge vs charge.

To test this, I kept an eye on my battery thermals while plugged in and gaming. Once the battery normalized at its max temp for 5 minutes or so, I unplugged the Air -- sure enough, the battery temps went up by 2C and performance marginally so -- the Apple was actually throttling based on battery temps as well, though in normal use it would never do so.

[JacksonWelch]

Would be interesting to see if the CPU mod of the M1 has any impact against these scores.

[ArcticBeats]

Would be interesting to include some benchmarks for music production, as Music Production Software (DAW) is heavily used on Macs. Additionally it can get very CPU heavy depending on the plugins you use.

[forgiv]

I'm a senior software developer that works on a few different kinds of projects. I bought the M2 Air as a portable machine that I could use for meetings and small amounts of web dev work while I'm out of the house. Honestly, it's been fantastic and has taken over as my primary machine for web projects. I use the base spec and it's honestly good enough as long as you don't use virtual machines or a jetbrains IDE. But if you use docker and vs code, then the base spec is more than good enough.

[iZian]

interestingly, late to the game I know sorry, but on a windows / intel machine: when you run on battery you don't get full performance by default do you? I mean; aren't the windows default to use a lower power setting on battery and doesn't the CPU throttle because it can't actually draw the wattage it would require from the battery without toasting it?
I mention this because the M1 / M2 machines: the default (I got one just the other day, first mac) seemed to be that "Low Power Mode" was set to never. As in; the CPU can run un-capped at full power on the battery, and that the OS does nothing to mitigate battery usage other than display / screen settings.
I noticed a dramatic difference when I ran my own tests using Low Power Mode on an M2 MacBook Air. The CPU and GPU seemed to be throttled / more work seemed to be delegated to efficiency cores, and some background tasks didn't bother to run like some indexing of images etc.

I wonder what the results would look like if these tests were run against the M1/M2 with LPM enabled on battery mode. It seemed like I'd be getting an extra 25-50% battery life depending on what I was doing.
But I didn't have enough time to sit there and let the thing drain in low power mode... I gave up once it got down to 65% because I didn't have the willpower to leave it running for a day solid