Apple MacBook Air M2 review - The faster 10-core GPU isn't worth it

We already knew that the new MacBook Air M2 has become significantly more expensive. With a well-equipped model as we have here (MSRP 2,309 Euros, ~$2,380), you already get into the range of a MacBook Pro 14, which is clearly superior in many areas (performance, Mini-LED display, connections, speakers). The basic model of the MBP 14 with 16 GB of RAM and a 512 GB SSD is currently available on the market for around 1,950 to 2,000 Euros (~$2,010-$2,061), and the model with a 1-TB SSD is available for about 2,300 Euros (~$2,371), which is the same price as our test unit of the Air M2. Of course, the Air targets a different audience with the better mobility, but we also include the Pro in our comparison charts.

• Apple MacBook Pro 14 basic version review (M1 Pro with 8 CPU cores and 14 GPU cores)
• Apple MacBook Pro 13 M2 (10 GPU cores, 16 GB RAM)
• Apple M2 SoC efficiency analysis

With the new MacBook Air M2 in the Midnight color, we determined that while the dark blue color looks very good, it is also very susceptible to dirt and fingerprints. The current test unit in the also new Starlight color version (silver with a slight golden hue) is significantly less susceptible in these regards and corresponds more to the known versions in Silver and Space Gray.

In our review of the basic version, we already mentioned that although the new MacBook Air M2 is very slim, it still has a larger footprint than the direct competitors such as the Dell XPS 13 Plus or the Asus ZenBook S 13. If we also compare it to the MacBook Pro 14, the Pro subjectively appears significantly bulkier and is, of course, also thicker and about 400 grams (~0.9 lb.) heavier. However, in terms of pure footprint, the difference isn't that large. 

We refer you to our review of the basic version for information on the display. In the comments section, we were asked about temporal dithering (it is used to create the illusion of more color depth). So, we examined the IPS panel of the Air M2 with a microscope and created a slow-motion recording with 240 images per second. While we were able to detect temporal dithering, for example, in the current MacBook Pro 16 with the Mini-LED panel in some particular gray color tones, this was neither the case in the old MacBook Air M1 nor in the new MacBook Air M2.

Since the processor module of the new MacBook Air always remains the same regardless of the configuration (with a total of 8 cores consisting of 4 Performance and 4 Efficiency cores), we can be brief here: The performance is basically identical. In the Cinebench loop, we see a slight advantage to the basic model after several runs, and in the final runs, the high-end Air is also minimally faster. However, this could also be explained with normal differences between the chips themselves. In practice, the difference is not noticeable.

Apple was able to increase both the single and multi-core performance due to the higher clock speed compared to the M1 processor, but you also pay for it with a higher power consumption. We have already shown in a separate article that the M2 therefore also doesn't run as efficiently as its predecessor in terms of the processor performance. 

However, the new M2 chip produces more than sufficient performance for everyday work overall, even in combination with the passive cooling of the MacBook Air. In fact, the new Air M2 is almost able to compete with the 8-core M1 Pro of the MacBook Pro 14, at least in the short multi-core benchmarks, although the performance of the Pro 14 remains completely stable without dropping. 

In terms of the subjective performance, we initially don't notice any difference to the basic model of the MacBook Air M2, especially in everyday tasks such as writing mails, playing videos, or surfing the Internet. However, in some benchmarks, the larger storage and also the faster SSD become noticeable. Basically, emulated programs also run without any problems via Rosetta 2, but the native apps for Apple chips are generally faster and more efficient on resources. A good example are the Adobe Photoshop and Premiere Pro apps that all run acceptably as Intel versions but are significantly slower than the native apps. The PugetBench results are therefore also only of limited use, but they still show the advantage of the increased working memory and faster SSD compared to the basic model of the MacBook Air M2.

Therefore, there is only a real advantage during short benchmarks. In everyday operation, the 8-core GPU is already more than sufficient. So there is no real reason to pay up the additional premium for the 10-core GPU in the new MacBook Air M2. However, we also want to make a brief comparison with the larger MacBook Pro 14 here, with even the 14-core GPU of the basic model being at least 30-40% faster and also remaining constant at this performance level.

The more expensive MacBook Air M2 is, of course, also passively cooled, so it remains completely silent during operation. We were also unable to detect any other electronic noises.

Since the passive cooling is identical in both models, there are also no real differences in the surface temperatures. There are no problems during everyday operation anyway, and even under high loads, Apple has a good handle on the case temperatures, so you can still run the MacBook Air M2 on top of your lap without any concerns.

During our stress test, we see basically the same behavior as in the basic model. In the beginning, the M2 is allowed to briefly consume about 33 watts, before the consumption drops rapidly to finally also level out more quickly at a slightly lower package power of less than 9 watts. This is also easy to see in the chart below, which compares the package power, CPU power, and GPU power of the two MacBook Air M2 models in our stress test with each other.

We already saw in the test of the basic model that the new speaker system of the MacBook Air M2 is again very good but doesn't represent a significant improvement compared to the previous MacBook Air M1. The MacBook Pro 14 with the additional modules is again better here and is only surpassed by the even larger MacBook Pro 16.

In the more-expensive models of the MacBook Air M2, the new dual USB-C charger is also included. This should be completely sufficient for most users, and the second USB-C connection can also be helpful in many situations. Optionally (and at no additional cost), you can also switch to the more powerful 67-Watt charger, which has a positive effect particularly on the recharging time, since the new MacBook Air M2 will then also support faster charging.

Due to the slightly higher power consumption, the measured battery life also turns out slightly worse than was the case with the basic model. However, the runtimes in the WLAN test and when playing videos are still very good if the display brightness is adjusted to 150 cd/m². At full brightness, the battery life is significantly shorter. 

The comparison with the MacBook Pro 14 is interesting again here, since although the runtime is about one hour shorter at 150 cd/m², the MacBook Pro 14 lasts even slightly longer when adjusted to the full SDR brightness (which is comparable to that of the new MacBook Air M2). While the new Air M2 doesn't quite last for a full 6 hours in the WLAN test, the MacBook Pro 14 is able to last for 6.5 hours. 

The MacBook Air M2 can be configured in the Apple Store and has a price starting from $1,199. Our test configuration is available for $1,899. Amazon also lists a version for $1,499.