M2 MacBook Air - Drama-Free First Impressions - Krazy Ken’s Tech Misadventures

You can support this channel on Patreon! Link below…
Some "drama" about the new M2 MacBook Air might have some merit to it, but let's set that aside and just have a good time with some raw reactions and first impressions!

👆Patrons got early access to this episode. They also got behind-the-scenes access, Discord VIP access, and their names in the credits! Pledge now to get those awesome perks, and help fund the future of the Computer Clan! Thanks for your support.

The Carina nebula wallpaper was captured with the James Webb Telescope and produced by NASA and the Space Telescope Science Institute (STScI).

Computer Clan is an Amazon Associate. Amazon links are powered by the Amazon Associates Program.

0:00 - Introduction
0:45 - Unboxing
2:39 - Color, Design, and Comparison
3:48 - Inside the Mac
5:19 - Temperature Test (Real-world Apps)
9:05 - AI Software Speed Test
10:42 - SSD Speed
11:48 - Geekbench (M2 vs. Original Intel Air)
12:25 - Final Cut Pro Speed Test (M2 vs. Original)
13:14 - A Little Teaser…

Correction: 12:01 The metrics between Geekbench 3 and 5 are different so these numbers really don't count. However, newer versions of Geekbench do not run on Mac OS X 10.6, so I could not test properly. Feel free to ignore these numbers. But ultimately, it just further proves that benchmarking (in most contexts) means nothing—especially when it's overdone and out of context.

#KrazyKen #ComputerClan

[ComputerClan]

Enjoy the new episode! And stay tuned for the 6, 000 Chrome tabs challenge—dropping on Thursday! 🔔 Also, big thanks to Linode for sponsoring today's episode!

[sotosoul]

Other viewers have commented on the different scale of Geekbench 3 and 5 scales. To bring everything to Geekbench 5 scale, I've calculated the following:
1st Gen MacBook Air Geekbench 5 adjusted scores (single/multi): 198 / 347
Therefore, the M2 MacBook Air is about *9, 5* × faster in single core and *25, 5* × faster in multi-core.

[friedgpu]

You're wrong when comparing the Geekbench scores because the score calibration differs between versions. Geekbench 3 uses a Mac mini (Mid 2011) with an Intel Core i5-2520M with a baseline of 2500 and Geekbench 5 uses a Intel Core i3-8100 with a baseline of 1000.

[jonglass]

I remember watching Steve Jobs pull out the first Air from an inter-office envelope, and jonesing so badly over it! I never bought myself an Air, but bought two for my family, and went with the MBP. but I still love the idea of the Air being so thin... I kind of miss the wedge shape of the new one, but it seems a worthy successor...

[exDeveloper]

I wanted to remind you that Geekbench 3 and Geekbench 5 measures the scores in extremely different ways and they do not represent 2X difference…

[markusTegelane]

I feel like 6000 tabs in Chrome is the ultimate test for the robustness of memory management on an operating system. Also a test of the hardware (including RAM, CPU, and storage) of course ;)

[lucymorrison]

I got my M2 Air in Starlight a few days ago and I'm smitten. It's gorgeous, it's fast, it's silent... From my 2015 Air it's such a massive upgrade.

[polypolyman]

Makes sense that a hardware-accelerated encode would hardly use the CPU - the CPU usage will only measure the actual cores themselves, not the encoders, so it won't go up if the encoders are busy. The amount you did see was likely just feeding data in and out (not trivial, but very easy compared to other compute tasks)

[GareWorks]

I wish we got a new Mac Mini too. Hopefully soon, because I'll probably buy an M2 Mac Mini. The Mac Studio is just way too expensive as an alternative, especially if you want more storage or RAM.
I do wish the M2 MacBook Airs came in the colours they did for the M1 iMac. The leaks were pretty consistent about that, so I wonder if perhaps they decided to drop the large array of colours for cost and/or supply reasons, and with less SKUs, they decided to stick closer to the iPhone colours instead of only doing some of the iMac colours.
I still wish Apple were repair-friendly but they're not, and they still make some maddening design and engineering choices. It beats dealing with Windows (pretty low bar), but it doesn't beat being able to build and repair everything yourself.

[itsAlphaFox]

Geekbench 5 has completely different scaling from previous Geekbench versions

[PearComputingDevices]

I love that wallpaper, it would go very well with dark mode. Beautiful!!!

[Dsjk]

An infrared camera would be cool for seeing what areas of the computers heat up during tests.

[lbsiuk]

The CPU usage will not be that high whilst you're encoding because you'll be using the hardware encoders built into the GPU instead of the CPU.

[NicolaMastrandrea]

Drama free review. We really need that!

[tipoomaster]

The encode is just hitting the encoder and the GPU compute, hence why it never spun up the performance CPU cores.

[davidbillberg]

2:27 You need the 19 dollar polishing cloth for it to be completely clean 😉

[ytrewq]

We love you too man! Keep up the hard work and have a great day!

[zephyr]

Was your test on first boot? If so, spotlight indexing usually hogs a good bit of the CPU until done.

[harrisonxXx]

Design wise - i would LOVE the M2 Air... but performance and price wise... i love my 16" M1 Pro much more!

[kassiepovinelli]

Having checked it out in-store, the notch is just too much for the screen size. The 16" MBP manages its notch much more gracefully.