MacBook Pro 2021 benchmarks — how fast are M1 Pro and M1 Max?

The MacBook Pro 2021 (14-inch)
(Image credit: Henry T. Casey)

Apple revealed its M1 Pro and M1 Max chips when it unveiled the new MacBook Pro 2021 (14-inch) and MacBook Pro 2021 (16-inch) laptops during the Apple Unleashed event

The company boasted about the power of the Apple silicon processors and made direct comparisons to PC laptops, to the extent that even non-Apple fans took notice of the raw power these new M1 processors supposedly possessed. On paper, the M1 Pro and M1 Max appear to be game-changers for the MacBook Pro line.

We now have numbers that prove the M1 processors are as powerful as we were led to believe. 

We’ve pitted the MacBook Pro 2021 14-inch (M1 Pro 10-core GPU with 32GB of RAM) and the MacBook Pro 2021 16-inch (M1 Max 32-core GPU with 64GB of RAM) against three comparable laptops: the Asus ProArt StudioBook 16, HP ZBook Fury 17.3-Inch G8, and Microsoft Surface Laptop Studio. Both the ProArt StudioBook and Surface Laptop Studio have 32GB of RAM, while the ZBook Fury has 64GB.

For good measure, we also compared it to last year's MacBook Air M1 and MacBook Pro M1 to give you a sense of how the new M1 chips stack against their predecessors.

MacBook M1 Pro and M1 Max: Performance benchmarks 

On the Geekbench 5.4 multi-core test, the M1 Pro and M1 Max scored 12,477 and 12,683, respectively. These scores obliterate the Microsoft Surface Laptop Studio’s 5,820 and handily surpass the Asus ProArt StudioBook 16’s score of 9,158 and the HP ZBook Fury’s 9,716. The M1 MacBook Air (5,962) and M1 MacBook Pro (5,925) can’t compete, either. 

Swipe to scroll horizontally
Header Cell - Column 0 Geekbench 5.4Handbreak video transcodingPugetBench Photoshop
MacBook Pro 14-inch M1 Pro12,4774:51806
MacBook Pro 16-inch M1 Max12,6834:48877
Asus ProArt StudioBook 169,1586:06DNR
HP ZBook Fury G89,7166:271,077
Microsoft Surface Laptop Studio5,82011:24DNR
MacBook Air M15,9629:15653
MacBook Pro M15,9257:44649

The new laptops nab another win on our Handbrake video transcoding test. They convert a 4K film to 1080p at 4:51 and 4:48, respectively; this is nearly half the 11:24 taken by the Surface Laptop Studio. Both the M1 Pro and M1 Max MacBooks also beat the Asus and HP laptops in this category, but not by so much. They also outshine the MacBook Air M1’s 9:15 and the MacBook Pro M1’s 7:44.

Finally, we have the sole performance category where the new M1 Pro/Max laptops lose. On the PugetBench Photoshop test, the M1 Pro (806) and M1 Max (877) scored less than the ZBook Fury (1017). However, their respective times — 4:54 for the M1 Pro and 4:44 for the M1 Max — are comparable with the ZBook Fury’s 4:56. We don’t have Photoshop numbers for the Surface Laptop Studio or Asus ProArt StudioBook 16. However, both new laptops surpass the MacBook Air M1 (653) and MacBook Pro M1 (649).

MacBook M1 Pro and M1 Max: Gaming benchmarks 

Gaming is one of the areas we expected the 2021 MacBook Pros to really shine. However, our tests reveal some surprising numbers. Sid Meier’s Civilization VI runs at 46 fps on the M1 Pro (at 1512x982p resolution) and 46 fps on the M1 Max (1728x1080). These numbers are marginally better than the 37 and 38 fps reached on the M1 MacBook Air and M1 MacBook Pro, respectively.

Swipe to scroll horizontally
Header Cell - Column 0 Civilization IV: Gathering Storm (fps)Resolution
MacBook Pro 14-inch M1 Pro461512 x 982
MacBook Pro 16-inch M1 Max461728 x 1080
Asus ProArt StudioBook 16921920 x 1080
HP ZBook Fury G8831920 x 1080
Microsoft Surface Laptop Studio661920 x 1080
MacBook Air M1371440 x 900
MacBook Pro M1381440 x 900

The new MacBooks cannot touch the whopping 92 fps achieved on the ProArt Studio Book 16 (1080p). Both the Surface Laptop Studio (66 fps) and ZBook Fury (83) also performed better. As I've said before, the MacBook Pro M1 Max is an absolute beast — and a complete waste for gamers.

MacBook M1 Pro and M1 Max: Battery life benchmarks 

Battery life is jaw-dropping compared to the competition. The M1 Pro (14:08) and M1 Max (15:31) MacBooks lasted far longer than the ProArt StudioBook (6:16) and ZBook Fury (7:09). They also bested the Surface Laptop Studio’s best time of 12:03. Like the MacBook Air M1 (14:41) and MacBook Pro M1 (16:32) before them, the new 2021 M1-powered MacBook Pros have exceptional all-day battery life. 

Swipe to scroll horizontally
Header Cell - Column 0 Battery life
MacBook Pro 14-inch M1 Pro14:08
MacBook Pro 16-inch M1 Max15:31
Asus ProArt StudioBook 166:16
HP ZBook Fury G87:09
Microsoft Surface Laptop Studio10:42 (120Hz)
MacBook Air M114:41
MacBook Pro M116:32

The M1 Pro and M1 Max-powered MacBook Pros are serious business, as revealed by our tests. Apple wasn’t joking when it said the new M1 chips are the company’s most powerful processors. Gaming performance is disappointing, but by every other measure, the new M1s outstrip the competition.

Tony Polanco
Computing Writer

Tony is a computing writer at Tom’s Guide covering laptops, tablets, Windows, and iOS. During his off-hours, Tony enjoys reading comic books, playing video games, reading speculative fiction novels, and spending too much time on X/Twitter. His non-nerdy pursuits involve attending Hard Rock/Heavy Metal concerts and going to NYC bars with friends and colleagues. His work has appeared in publications such as Laptop Mag, PC Mag, and various independent gaming sites.

  • therealcaptaincrunch
    The author has made a rookie mistake or is being intellectual dishonest with his readers. The author fails to mention that "Sid Meier’s Civilization VI" is a NON-NATIVE application written for the INTEL x86 Mac and has to be emulated on the M1 Mac in Rosetta 2's translation layer, which invalidates his claim that M1 Macs are not good for gaming. Try comparing two NATIVE games or applications and you'll see that it easily matches the best high-end mobile graphics chips, if not beating them with far lower power consumption.

    Also, in performance per/watt, which is what matters in a laptop. The M1, M1 Pro and M1 Max destroy the INTEL and AMD chips.
    Reply
  • rgd1101
    Do you have any suggestion for native games that the author can use?
    Reply
  • McD
    What's with those Geekbench scores for the original M1s? The Geekbench Browser site puts them at 7600 not under 6000! It's hard to trust the other scores after that.

    Also, whilst the x86 vendor PugetBench is as biased as Intel AVX2 CPU renderer Cinebench, where are these figures on their site? Why are there only Premiere figures posted? I also see that screen refresh impacts the score so can TG disclose they had the MBP screen set to 120Hz?

    Can also TG disclose if they were using the VideoToolbox when encoding in Handbrake as you can't really show off a SoC if you ignore the fast bits.
    Reply
  • therealcaptaincrunch
    rgd1101 said:
    Do you have any suggestion for native games that the author can use?


    The point is that these tests were run using Mac x86 games being emulated in Rosetta, something the author fails to mention. Without telling his readers this, many will conclude that the M1 not good for gaming with native software instead of in emulation. I’ve seen articles like this repeated several times and people unfamiliar with Macs and the often poor x86 ports are simply coming to the wrong conclusions.

    There are at this point few native games for the M1 on sale because many developers are still transitioning from x86 Mac to M1 Mac.

    But one thing is different now that was not in the Mac market before, the plethora of Native iOS games for iPhone and iPad that already run native on the M1, like Civilization IV. The problem is that the Mac version in the AppStore is x86 but the iPad/iPhone version is native M1 ARM. The iPhone/iPad native versions will run on the M1 but developers have not allowed it yet, probably because the interface is designed for touch.

    The bottom line is that these gaming tests prove nothing about the capabilities of the GPU for gaming. Once the native versions come along we will see very different results.
    Reply
  • rgd1101
    good point. they shouldn't get it for game, at least not yet. not sure why they even benchmark them.
    Reply
  • McD
    rgd1101 said:
    Do you have any suggestion for native games that the author can use?
    https://www.macgamerhq.com/apple-m1/native-mac-m1-games/
    Whilst only a few are M1 native (though I think Baldur's Gate has just converted), it's odd that so many already support the Metal API (where the graphics work is done) but haven't been recompiled for ARM. I would say ARM is an inevitability in the PC world. Maybe the M1 Pro/Max is the excuse they need whilst getting even a minimal return on their investment.
    Reply
  • therealcaptaincrunch
    McD said:
    https://www.macgamerhq.com/apple-m1/native-mac-m1-games/
    Whilst only a few are M1 native (though I think Baldur's Gate has just converted), it's odd that so many already support the Metal API (where the graphics work is done) but haven't been recompiled for ARM. I would say ARM is an inevitability in the PC world. Maybe the M1 Pro/Max is the excuse they need whilst getting even a minimal return on their investment.

    The iOS games market is huge and we will definitely see many of those native games scaled up to the desktop and the economy of scale the Mac didn’t have before will now com into play. The very game they tested here already has a native iOS version, so it won’t be long before these misguided benchmarks are history.
    Reply
  • Byte99
    This review is, alas, meaningless. First, as mentioned by the commenters above, his gaming benchmark used a game that can't run natively on the Mac ( and he did this without informing the reader).

    Second, more generally, none of his benchmarks of the Macs vs. other mfrs. tell us anything, because we have no idea what he's comparing the Macs against.

    Take, for instance, the HP ZBook Fury 17 G8. That's a highly customizable workstation. Specifically, it can range from a low-end Core i5-11500H (6 cores) with Intel UHD integrated graphics, to an ultra-high-end* Xeon W-11955M (8 cores) with NVIDIA RTX A5000 GPU (16 GB DDR). Without knowing which configuration of ZBook he's using (the only spec we're given is that it has 64 GB RAM), all his comparisons relative to the ZBook tell us nothing.

    *With this CPU-GPU combo, 64 GB RAM, 2 TB HD, and a 4K display, it's $7,050, which is $2,750 more than an M1 Max with the same RAM and TB size ($4,300).
    https://www.hp.com/us-en/shop/ConfigureView?langId=-1&storeId=10151&catalogId=10051&catEntryId=3074457345620032820&urlLangId=&quantity=1
    And it's the exact same problem with the Asus ProArt StudioBook 16 and the Microsoft Surface Laptop Studio. All he gives us is the RAM. It's so frustrating when reviews lack the basic information needed to make them meaningful.
    Reply
  • vekspec
    I mostly do Adobe CC (Lightroom Classic, Photoshop, etc) and after reading some of the reviews on Adobe for the new M1s, I decided to bench my desktop and got a PS score of 1220. guess I'll just transition from my 2019 MBP 16 to my PC desktop ;)
    Reply
  • therealcaptaincrunch
    vekspec said:
    I mostly do Adobe CC (Lightroom Classic, Photoshop, etc) and after reading some of the reviews on Adobe for the new M1s, I decided to bench my desktop and got a PS score of 1220. guess I'll just transition from my 2019 MBP 16 to my PC desktop ;)

    What reviews? What are you even saying?
    Reply