Apple Silicon M1: benchmarks and review for Mac Mini, MacBook Air and MBP 13 M1 and comparison with other brands

[bill-p]

In their presentation, apple said M1 powered laptops were faster than 98% of all laptops being sold at present. In other words, the top 2% of laptops being sold are still faster by some measure, but every faster laptop operates with a much higher TDP.

Apple is hitting roughly 1.5x the performance per watt of AMD’s most efficient parts (4800U in 15W configuration), and roughly 2-3x (depending on configured TDP) the performance per watt of Intel’s most efficient parts (i7-1165G7). These efficiency advantages will definitely result in advantages for Apple as they scale to higher core counts. Increasing clock speeds increases power consumption very non-linearly, but increasing core counts generally scales near-linearly, or even sub-linearly if you clock the cores slower. A future 16” MBP with Apple Silicon may still be outperformed by high power and high core count desktops, but I’m confident Apple can be best in class for performance at every power budget by a comfortable margin.

Yeah, I have no doubt Apple will prevail with the 16" and the iMac, Mac Pro as well.

But the problem at this point is... it'll be Apple against itself when those machines come.

The reception of the M1 has been generally positive. But... imagine the reception when Apple introduces a 16" MacBook Pro that looks essentially the same as last gen, with basically no single-core improvement over the 13", and with only more CPU cores and more GPU cores, and with less battery life (likely 15 hours like the Air).

And then imagine the reception when they introduce iMac and Mac Pro that are basically... even more cores.

So yeah, that's the problem. Apple set its own bar, and I can't help but see that the rest of the lineup may not be as impressive.

 

[Henk van Ess]

Comparison made against latest‑generation high‑performance notebooks commercially available at the time of testing.

Less generic , but still not precise enough, are the 20+ footnotes on any product page of M1 models, which reveal some info, like 2.8x faster processing power (see 2) or 5x faster graphics (see 8)

1. 
   
2. Testing conducted by Apple in October 2020 using preproduction 13-inch MacBook Pro systems with Apple M1 chip, as well as production 1.7GHz quad-core Intel Core i7-based 13-inch MacBook Pro systems, all configured with 16GB RAM and 2TB SSD. Open source project built with prerelease Xcode 12.2 with Apple Clang 12.0.0, Ninja 1.10.0.git, and CMake 3.16.5. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro.
3. 
   
4. Testing conducted by Apple in October 2020 using preproduction 13-inch MacBook Pro systems with Apple M1 chip, as well as production 1.7GHz quad-core Intel Core i7-based 13-inch MacBook Pro systems, all configured with 16GB RAM and 2TB SSD. Tested with prerelease Final Cut Pro 10.5 using a 55-second clip with 4K Apple ProRes RAW media, at 4096x2160 resolution and 59.94 frames per second, transcoded to Apple ProRes 422. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro.
5. Testing conducted by Apple in October 2020 using preproduction 13-inch MacBook Pro systems with Apple M1 chip, as well as production 1.7GHz quad-core Intel Core i7-based 13-inch MacBook Pro systems, all configured with 16GB RAM and 2TB SSD. Tested with prerelease Affinity Photo 1.9.0.199 using the built-in benchmark version 1900. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro.
6. Testing conducted by Apple in October 2020 using preproduction 13-inch MacBook Pro systems with Apple M1 chip, as well as production 1.7GHz quad-core Intel Core i7-based 13-inch MacBook Pro systems, all configured with 16GB RAM and 2TB SSD. Tested with prerelease Logic Pro 10.6.0 with project consisting of multiple tracks, each with an Amp Designer plug-in instance applied. Individual tracks were added during playback until CPU became overloaded. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro.
7. Testing conducted by Apple in October 2020 using preproduction 13-inch MacBook Pro systems with Apple M1 chip and 16GB of RAM using select industry-standard benchmarks. Comparison made against the highest-performing integrated GPUs for notebooks and desktops commercially available at the time of testing. Integrated GPU is defined as a GPU located on a monolithic silicon die along with a CPU and memory controller, behind a unified memory subsystem. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro.
8. Testing conducted by Apple in October 2020 using preproduction 13-inch MacBook Pro systems with Apple M1 chip, as well as production 1.7GHz quad-core Intel Core i7-based 13-inch MacBook Pro systems with Intel Iris Plus Graphics 645, all configured with 16GB RAM and 2TB SSD. Tested with prerelease Final Cut Pro 10.5 using a 10-second project with Apple ProRes 422 video at 3840x2160 resolution and 30 frames per second. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro

 

[rawCpoppa]

Yeah, I have no doubt Apple will prevail with the 16" and the iMac, Mac Pro as well.

But the problem at this point is... it'll be Apple against itself when those machines come.

The reception of the M1 has been generally positive. But... imagine the reception when Apple introduces a 16" MacBook Pro that looks essentially the same as last gen, with basically no single-core improvement over the 13", and with only more CPU cores and more GPU cores, and with less battery life (likely 15 hours like the Air).

And then imagine the reception when they introduce iMac and Mac Pro that are basically... even more cores.

So yeah, that's the problem. Apple set its own bar, and I can't help but see that the rest of the lineup may not be as impressive.

I think Apple knows what it’s doing and will bring compelling products to the market in 16 inch MBP or iMac configs.

 

[M1 Processor]

In their presentation, apple said M1 powered laptops were faster than 98% of all laptops being sold at present. In other words, the top 2% of laptops being sold are still faster by some measure, but every faster laptop operates with a much higher TDP.

Apple is hitting roughly 1.5x the performance per watt of AMD’s most efficient parts (4800U in 15W configuration), and roughly 2-3x the performance per watt of Intel’s most efficient parts (i7-1165G7). These efficiency advantages will definitely result in advantages for Apple as they scale to higher core counts. Increasing clock speeds increases power consumption very non-linearly, but increasing core counts generally scales near-linearly, or even sub-linearly if you clock the cores slower. A future 16” MBP with Apple Silicon may still be outperformed by high power and high core count desktops, but I’m confident Apple can be best in class for performance at every power budget by a comfortable margin.

I never said that they said they said there were the fastest. In addition I said its the most efficient silicon, but it is on 5nm after all and the Zen 2 core is a bit old in the tooth. Of course there are additional trade offs not being mentioned. For example, its I/O is inferior to Intel/AMD’s offerings and bringing it up to par probably would have eaten into its efficiency lead. Of course its µarch is more than a year newer than AMD’s and its process node of a full 2 generations newer. It seems x86 is more competitive than some people claim, like Apple probably could have made as equally as good an x86 core if they had the IP.

 

[sultanq]

I am doubtful they are comparing AMD for two reasons.

1. Everything that they said up into this point compared Intel, and can’t directly compared with AMD since AMD silicon is not available for macOS. Geekbench scores are a bit higher on macOS than windows btw.

2. Their benchmarks don’t line up with what we know about chips such as the 4800u. As a point of reference the 4800u scores about 10000 in R23 compared to 7600 in the M1. Of course the 4800u is probably its 25w tpd config, vs 15 for the MBP.

One thing to note is that Maxon has had decades to optimize their rendering engine for x86, and several years optimizing for modern variations of recent Intel architectures, whereas their rendering engine is just freshly ported to AArch64. I'd be interested at seeing comparisons of applications that have had a longer presence on ARM. For instance, Geekbench has a much wider variety of workloads, and its implementations of various algorithms have undergone years of optimizations for both architectures, so I'd consider it a pretty good measure of peak performance. However, Geekbench is bursty, and is thus not representative of what can be achived for sustained load within a particular TDP.

 

[M1 Processor]

One thing to note is that Maxon has had decades to optimize their rendering engine for x86, and several years optimizing for modern variations of recent Intel architectures, whereas their rendering engine is just freshly ported to AArch64. I'd be interested at seeing comparisons of applications that have had a longer presence on ARM.

Historically Cinebench is a pretty good benchmark and it seems to line up very well with what Apple has said. It has shown itself to be much more reliable than Geekbench for sure (i.e geekbench failed to show the difference between the Air and Pro). Apple said its the fastest low power core, and its roughly that on Cinebench with Tiger Lake not really being that available. The thing with the M1 is just it just has less cores work with than AMD. Its not a bad thing, its basically the A14x (which is probably the reason behind the limited I/O), so I doubt that Apple tried really hard here.

Of course it remained to be seen how Apple will fair against next generation Threadrippers, but its a good time to be a computer enthusiast

 

[EugW]

Still very impressive! Especially considering it is passively cooled 10watt CPU. Not to far off from the base config 16" Pro in multicore and still beats it in single (6912/1113).

Again, that 10 Watts is just a number somebody here made up.

 

[sultanq]

But... imagine the reception when Apple introduces a 16" MacBook Pro that looks essentially the same as last gen, with basically no single-core improvement over the 13", and with only more CPU cores and more GPU cores, and with less battery life (likely 15 hours like the Air).

Battery life at full load would be reduced, but usually battery life is reported under constant workloads (eg. web browsing) rather than workloads that scale with CPU power. Unused cores can be clocked down or turned off altogether when doing things light workloads like battery life tests. The battery capacity to CPU current draw ratio for a 16" model should be better than a 13" under a fixed workload, so I'd expect the battery life of the 16" model with Apple Silicon to be longer than the 16" under things like web browsing and video playback tests used to measure battery life.

 

[M1 Processor]

Again, that 10 Watts is just a number somebody here made up.

“M1 delivers significantly higher performance at every power level when compared with the very latest PC laptop chip. At just 10 watts (the thermal envelope of a MacBook Air), M1 delivers up to 2x the CPU performance of the PC chip. And M1 can match the peak performance of the PC chip while using just a quarter of the power.”

 

[sultanq]

Historically Cinebench is a pretty good benchmark and it seems to line up very well with what Apple has said. It has shown itself to be much more reliable than Geekbench for sure (i.e geekbench failed to show the difference between the Air and Pro). Apple said its the fastest low power core, and its roughly that on Cinebench with Tiger Lake not really being that available. The thing with the M1 is just it just has less cores work with than AMD. Its not a bad thing, its basically the A14x (which is probably the reason behind the limited I/O), so I doubt that Apple tried really hard here.

Of course it remained to be seen how Apple will fair against next generation Threadrippers, but its a good time to be a computer enthusiast

Geekbench is bursty, and the CPU is idle for a fair percentage of the time even during the multi-core phase of the test. Thus, Geekbench is a measure of peak performance, rather than sustained performance under the current thermal envelope. Most likely both the MBP and MBA have their CPUs drawing something like 20+ watts during the multi-core portion of the test, but the average power draw remains within the MBA's limits, and the duration of the multi-core test isn't enough to max out the heat capacity of the heatsink. It's worth noting that for many users (myself included), most performance intensive tasks are short in duration, so peak performance for short bursts is important, and for that the MBP and MBA are equal.

I'd love to see Geekbench implement an alternate mode where it runs the multi-core tests back-to-back repeatedly for an extended period, akin to the thermal stress testing of Cinebench.

 

[Zackmd1]

Again, that 10 Watts is just a number somebody here made up.

Um..... The current intel air has a TDP around 10 watts and Apple themselves admitted to a 10 watt TDP for the M1 in the Air during the keynote. So no, 10 watts is a real number and not one that was made up...

 

[EugW]

Um..... The current intel air has a TDP around 10 watts and Apple themselves admitted to a 10 watt TDP for the M1 in the Air during the keynote. So no, 10 watts is a real number and not one that was made up...

Hmm... If so, I stand corrected.

EDIT:

I see that blurb now. That's an interesting way to word it. It sounds like they are saying that it's a potentially higher TDP chip, but in that 10 W thermal envelope it will do xyz performance.

That's true, but the real TDP of the chip would be higher, as is evidenced the performance difference with the MacBook Pro. In the 10W thermal envelope, it throttles and loses over 15% performance, meaning that isn't really its native TDP.

 

[M1 Processor]

Geekbench is bursty, and the CPU is idle for a fair percentage of the time even during the multi-core phase of the test. Thus, Geekbench is a measure of peak performance, rather than sustained performance under the current thermal envelope. Most likely both the MBP and MBA have their CPUs drawing something like 20+ watts during the multi-core portion of the test, but the average power draw remains within the MBA's limits, and the duration of the multi-core test isn't enough to max out the heat capacity of the heatsink.

I’m VERY skeptical of Geekbench, even though it appears to have gotten better. I will give an example: the Exynos 9820 thrashed the in Snapdragon 855/Kirin 980 in Geekbench, but lost in every other benchmark against the Snapdragon 855, and not only lost to the 855, but often underperformed the 845. When I mean lost, it lost by a lot. I have numerous examples. I trust Cinebench much more than I do Geekbench.

 

[bill-p]

Hmm... If so, I stand corrected.

I see that post now. That's an interesting way to word it. It sounds like they are saying that it's a potentially higher TDP chip, but in that 10 W thermal envelope it will do xyz.

That's true, but the real TDP of the chip seems to be higher, as is evidenced the performance difference with the MacBook Pro. In the 10W thermal envelope, it throttles and loses over 15% performance, meaning that isn't really its proper TDP.

And you are making me wonder how many people bought the MacBook Air without knowing this...

Then again, can't blame them when all we had for most of last week were Geekbench scores that basically showed zero difference in performance between the Air and Pro.

 

[thingstoponder]

Yeah, I have no doubt Apple will prevail with the 16" and the iMac, Mac Pro as well.

But the problem at this point is... it'll be Apple against itself when those machines come.

The reception of the M1 has been generally positive. But... imagine the reception when Apple introduces a 16" MacBook Pro that looks essentially the same as last gen, with basically no single-core improvement over the 13", and with only more CPU cores and more GPU cores, and with less battery life (likely 15 hours like the Air).

And then imagine the reception when they introduce iMac and Mac Pro that are basically... even more cores.

So yeah, that's the problem. Apple set its own bar, and I can't help but see that the rest of the lineup may not be as impressive.

How is that bad though? People buy the 16” for more cores and a bigger GPU. Single core being the same is not an issue as it’s already a huge upgrade from previous machines. The Mac Pro was already outpaced by higher end iMacs in single core.

Of course the smallest device you can fit your world class CPU core is always gonna be most impressive. You could argue the iPhone makes the Macs look less impressive by comparison.

 

[thingstoponder]

I’m VERY skeptical of Geekbench, even though it appears to have gotten better. I will give an example: the Exynos 9820 thrashed the in Snapdragon 855/Kirin 980 in Geekbench, but lost in every other benchmark against the Snapdragon 855, and not only lost to the 855, but often underperformed the 845. When I mean lost, it lost by a lot. I have numerous examples. I trust Cinebench much more than I do Geekbench.

Cinebench is just a single test for a single video editing app, Cinema4d. AMD for example always uses it because it makes them look good. It’s not the be all end all benchmark either. It’s best to look at a wide range of tests, including geekbench.

 

[rawCpoppa]

And you are making me wonder how many people bought the MacBook Air without knowing this...

Then again, can't blame them when all we had for most of last week were Geekbench scores that basically showed zero difference in performance between the Air and Pro.

I think that’s on them. No geekbench score would convince me that the air would be equal or better than the pro for a cheaper price. Apple themselves mention the active cooling so the difference had to show up somewhere.

 

[bill-p]

How is that bad though? People buy the 16” for more cores and a bigger GPU. Single core being the same is not an issue as it’s already a huge upgrade from previous machines. The Mac Pro was already outpaced by higher end iMacs in single core.

Of course the smallest device you can fit your world class CPU core is always gonna be most impressive. You could argue the iPhone makes the Macs look less impressive by comparison.

The 16" 2019 is faster than 13" 2019 in both single core and multi core scenarios. Not only that, it has better battery life and better screen/keyboard as well. It was better in every way possible except for maybe size and weight.

And the 13" M1 MacBook is faster than A14 in iPhone 12 Pro in every way possible. The iPhone 12 Pro is only... portable, but that's about it. There's no way you can make an iPhone 12 Pro last 20 hours with the screen on all that time, but supposedly, that should be possible with the 13" Pro.

So there was always a natural progression in terms of performance scaling and value proposition.

I'm just not sure we'll see that with the next 16" compared to the 13" M1 if Apple isn't willing to sacrifice on efficiency. So there will be drawbacks and tradeoffs.

 

[EugW]

I’m VERY skeptical of Geekbench, even though it appears to have gotten better. I will give an example: the Exynos 9820 thrashed the in Snapdragon 855/Kirin 980 in Geekbench, but lost in every other benchmark against the Snapdragon 855, and not only lost to the 855, but often underperformed the 845. When I mean lost, it lost by a lot. I have numerous examples. I trust Cinebench much more than I do Geekbench.

Cinebench is a good measure of sustained Cinebench-type CPU workload, but it isn't really a good gauge of other stuff.

If you want to see if your fanless computer throttles though, it's a good bench. Or if you want to see how loud your fans get, it's a good bench.

And you are making me wonder how many people bought the MacBook Air without knowing this...

Then again, can't blame them when all we had for most of last week were Geekbench scores that basically showed zero difference in performance between the Air and Pro.

If they don't understand how Geekbench works, that's their problem. Geekbench is a good bench for what it is. It is a good general purpose measure, but it's lousy at measuring sustained workloads, because it doesn't last very long. OTOH, for bursty workloads, it's very good.

But that's OK, because in general, people who use Airs work in bursts. For example, rendering a complex multi-media webpage still only takes seconds, and then the computer just sits there doing nothing, or almost nothing. It's not as if it has to render that webpage continuously at full tilt for 20 minutes.

To oversimplify things, Geekbench tells you how fast the computer will feel when you're sitting at it and actively using it, whereas Cinebench tells how long that long render is going to take and how loud your fans are going to get.

 

[nikidimi]

Why do you think this is a valuable comparison, comparing a 4 core SoC with an 8 core SoC? Is it something to be proud of that it takes 6 to 8 x86 large cores and a lot more power to match the M1? I just don't understand what's going on here. Do all these people truly not understand the difference between a large and a small core?

Yes, we all want the M1X to be released soon. But until Apple releases its machines in that class, with 8 large cores, we are limited to the M1.

Valuable to whom? We know that Zen 2 single core performance is not very impressive and we know that they are no 10W x86 CPU. What's the point in discussing that? We can compare it to a server CPU if we find it interesting for some reason. I think a lot of people are taking this thread way too seriously. We are not the official judges of the "worlds best CPU" competition, so we can make unfair comparisons. And I think that's interesting for people that want to have a good overview of the whole CPU market and I think that's in line with the topic.

 

[Henk van Ess]

Worst first review ever

 

[Zackmd1]

And you are making me wonder how many people bought the MacBook Air without knowing this...

Then again, can't blame them when all we had for most of last week were Geekbench scores that basically showed zero difference in performance between the Air and Pro.

I knew full well what the difference was going to be when I canceled my initial pro order and went with the air instead. I prioritized the fact that the air is lighter and passively cooled over sustained performance in addition to being able to get more storage while still being cheaper then the base pro. Most of my workflow consists of burst operations anyways instead of sustained load so the air and pro should be pretty comparable for me. That’s not even considering the fact that the air is still faster then most of the current Intel MBP 13” configs.

It was heavily hinted throughout the keynote that the M1 targets 15watts as its max TDP.

 

[Henk van Ess]

This one is live right now

 

[Henk van Ess]

Unboxing

 

[vocalnick]

Worst first review ever

I would like to apologise on behalf of Australia.