"But the low-priced version is the low-priced version, isn't it?" Or "Is it just the keyboard of the lower model of the old model replaced?" maybe.
However, when I actually tested it, far from being surprisingly good, despite the fact that the MacBook Air was only 10,000 yen higher, the MacBook Pro was clearly superior in terms of actual performance. Speaking only of a simple Geekbench 5 score, both have good points … That's the result, but when I actually run the application, this is a big difference.
By the way, MacBook Air's Core-i5 model is 8GB memory / 512GB SSD for 148,000 yen, MacBook Pro's model with two Thunderbolt 3 ports has the same price as 256GB SSD. If you add 10,000 yen here, it will be the same 8GB memory / 512GB SSD configuration as MacBook Air.
The bottom line is that for higher performance applications, the MacBook Pro outperforms the price difference. Of course, the battery lasts an hour or so on the specs, or the wedge shape is good! Or, I only want to have gold! I do not deny such needs. Some may not miss the 110g lighter spec.
In fact, if you're just writing a document or doing something on the web, you probably don't see a big performance difference. However, when I actually moved it, there was an unexpected big difference.
A simple bench is a 10th generation Macbook Air that overwhelms the GPU, but
The MacBook Air has a 10th generation Core processor. This processor is equipped with an instruction set and additional circuits that accelerate machine learning processing, and various improvements and expansions such as LPDDR4X compatible and WiFi-6 compatible have been added. It's only natural that the two generations have evolved, so it's better, but modern systems aren't running at clock frequencies to specs. It is optimized for the cooling system and the surrounding temperature environment. And how many cores can move at the same time and at what speed. The tuning around that depends on the hardware design.
On the other hand, it is an 8th generation Intel Core with two Thunderbolt 3 ports MacBook Pro, but the same MacBook Pro also has a 15 watt TDP processor. In addition, the upper model with 4 Thunderbolt 3 ports equipped with Ice Lake is compatible with processors up to 28 watts, and you can also determine whether there are large intake slits on both sides of the bottom. Although it has the same name, the airflow design is actually different.
Now, with Turbo Boost, it can be said that the maximum is ~, but how much clock frequency actually works depends on the surrounding environment, and more specifically, it depends on the system design. How different is the performance from the MacBook Air that seems to have little airflow [actually, the installed processor is 10W]? Do you mind?
By the way, when you do Pro vs Air with GeekBench 5, single core score is about 950 to 1100, multi core is about 3850 to 2600. For GPU [Compute, Metal], it is about 7500 vs. 7800. The mystery is still the GPU score. In the 2019 Late MacBook Pro, the built-in Intel Iris Plus Graphics 654 scored around 5100. However, I was surprised that the score suddenly appeared at 7499, but when I confirmed it with the 2018 model at hand, the Intel Iris Plus Graphics 655 also scored 7,000 units. In other words, at least the GPU is the same as the Air.
However, these numbers can fluctuate, so please think about it roughly. It's not very important in comparing real applications.
So I tried running some apps.
Higher performance on lower MacBook Pro models in any situation
First, video editing software. Here, with Blackmagic Design's DaVinci Resolve 16 which can use almost the same functions as for professionals, it is a 16-minute video [Full HD / 30fps] that overlaps multiple videos and inserts a telop while applying transition with cross resolve. I wrote it down.
When exporting a video, this application does a great deal of the video processing on the GPU. The GPU is used not through Open CL but through the metal native to macOS, which is an advantageous condition for Ice Lake equipped with MacBook Air.
However, I compared it while confirming that I used the GPU properly and fully, but it took about 24 minutes for the MacBook Air and 19 minutes for the MacBook Pro.
Actually, the throughput was fairly close at the beginning of writing, but the speed difference became clear in about 1 minute. Perhaps, it may have been put in control to suppress heat generation under heavy load.
Next, of the RAW development software, I tried batch development of 100 EOS R RAW files [CR3] using Luminar 4, which is reputed to be "heavy" despite its high image quality, while also using AI automatic discrimination phenomenon processing. ..
Again, the GPU runs around as much as possible, but at the same time the CPU also runs, which is a fairly severe condition. Then, it took 49 hours on the MacBook Pro, but it took 1 hour and 16 minutes on the MacBook Air. When I ran it several times, I could confirm the shaking for several minutes, but the difference is overwhelming and the MacBook Air will not reverse even if there is some error.
The CPU processing is also "sustainable" by the 8th generation Pro.
Originally, we have not been able to lead with the new generation GPU that should be faster, so this result is even more so for applications where GPU processing is a big factor, so it is even more so for CPU processing.
When rendering with single core with Cinebench R20 which performs 3D rendering processing with CPU, while the single core of Geekbench 5 has a better score on MacBook Air, the result is almost the same [slightly Pro It was better] [Pro vs Air 397 vs 355].
Originally, multi-core is an area where MacBook Air is not good at, so the difference is widened to 1566 vs. 872 with Pro vs. Air. This is probably because the MacBook Air's multi-core performance is limited to keep it cool.
Like the benchmark results, the one that is most affected is the RAW development process.
I introduced the test results of Luminar 4 earlier, but development software such as Camera Systems RAW from Adobe Systems or Digital Photo Professional from Canon does not use GPU at all [there are only ones at hand, so there may be others that use GPU]. Not]. In macOS Catalina, a function to display how much GPU each application uses is added to the activity monitor, so check how much software you are using GPU May be good.
Other differences between Pro and Air
Actually I did not think that the Thunderbolt 3 port would evaluate two models, but when I seriously evaluate it and when the price is only 10,000 yen different from the MacBook Air, it becomes a little heavy But I want to recommend Pro, and I feel that I would choose this. When the current MacBook Air was announced, the MacBook Pro was still expensive, but in addition to the change in the exchange rate being set and the SSD unit price falling, it changed to a fairly bargain model I think
By the way, if you look closely, there are the following differences between the MacBook Pro and MacBook Air.
- The maximum brightness of the display is brighter with Pro [500nits vs. 400nits with Air]
- Pro has wider color gamut of display [Air is sRGB compared to Display-P3]
- Pro with wider trackpad
- Cooling performance is higher for Pro [Air is 10 watt design for Pro that supports up to 28 watt processor]
- The sound quality and playback band of the speaker are adjusted for Pro [there is little distortion in the high range and the middle range is thick, but the sound pressure is slightly higher for Air]
- Presence or absence of Touch Bar [Some people say that it is not necessary, but it may be convenient for applications that are used to it]
Cost-effective Pro low-cost version
So, in terms of cost-effectiveness, the honest feeling is that the Macbook Pro is better than the MacBook Air.
Of course, a MacBook Pro with four Thunderbolt 3 ports with a 28-watt Ice Lake will do much better than that. Test results and reviews of high-end models have already appeared in the United States. We will send you a test about it tomorrow or the day after tomorrow, so if you are wondering "what happens to the higher model?", Please look forward to it.