M1 Max MacBook Pro - A Programmer's Perspective

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
hello world noah here the wait is finally over almost a year after announcing apple silicon apple has shipped their first 14-inch and 16-inch macbook pros to feature their custom silicon along with a brand new design and i picked up a brand new 16-inch macbook pro to replace my 2018 15-inch macbook pro back when the first apple silicon max were launched i made a video looking at the initial state of the developer ecosystem on apple silicon and i found that while things looked pretty good thanks especially to the rosetta 2 translation layer there were some holes in many people's workflows so in this video we're going to take another look one year later and see if things have improved we'll also run some benchmarks because everyone loves to see numbers and i'll talk a bit about the new design before giving my overall opinion will i keep the new 16-inch macbook pro or go back to my trusty old 2018 15-inch macbook pro let's dive in [Music] first let's talk about the state of the developer ecosystem on apple silicon things were already pretty good when the first apple silicon max launched thanks in part to mac os being a unix-based operating system and also of course the rosetta ii translation layer that was able to run a lot of the x86 tools before they were ported over to arm now almost a year later things are looking even better the homebrew package manager is now natively supported i made a video a while ago explaining how to install homebrew natively on an apple silicon mac but now you can just copy and paste the command from their website they also provide pre-compiled binaries so instead of having to download the source and then build from source for every package you want to install you can just download a pre-compiled copy and install it directly so homebrew is a lot faster to use now we also have native support for docker that was a big sticking point when apple silicon first came out but now docker fully supports apple silicon on the stable branch there are also native versions of all of the jetbrains ides and you can now emulate android natively on arm there's also a lot of other applications that have gotten apple silicon support since the developers have had almost a year to recompile their apps i want to note that a lot of the improvements i just mentioned have actually been around for a while but i hadn't had an apple silicon mac since i tested the m1 macbook air when it first came out so all these improvements are new to me the only non-native things i've noticed in my personal workflow are the flutter cli and cocoapods but you can just install them as x86 binaries and then you can run them from the terminal even if it's running in arm mode and rosetta 2 will take care of things totally seamlessly cocoapods was actually a bit tricky to install but after doing some googling i found out that you could just add a prefix to any command to force it to run in x86 mode so that's just a good tip to have so things are looking pretty good but there is one major issue that i still see and that's virtualization if you need to run other operating systems on your mac then you may need to take some pause here first x86 is basically out of the question if you want to run windows 10 or windows 11 in x86 and play you know all of your games or run x86 applications you're not really going to have any luck doing that on apple silicon and you'd be much better off sticking with an intel mac and using bootcamp or some sort of virtualization software if you want to run arm versions of windows and linux you do have two choices utm and parallels utm is free and open source but it lacks hardware acceleration so the performance is not very good and the application itself is also a bit janky and then parallels will run very well but it's paid software and it costs eighty dollars a year there is also crossover which allows you to run windows applications on a mac but your mileage will vary depending on the application that you try to run and of course it is paid software so to sum up virtualization if you rely heavily on bootcamp you're not going to find it on apple silicon and if you rely heavily on virtual machines then the experience is not super great on apple silicon especially if you're trying to use x86 in those cases i would recommend sticking with an intel machine or perhaps buying a windows machine to use for those purposes and then keeping your mac machine for your mac stuff now it's time to talk about benchmarks you've probably already seen some benchmarks for the new machines so we'll talk briefly about the general benchmarks and then get into some of the developer specific ones i picked up a fully specced 16-inch macbook pro with 64 gigabytes of unified memory and this will be going up against my 2018 15 inch macbook pro with the core i9 the radeon pro 560x and 16 gigs of ram i also have some numbers from the m1 macbook air that i benchmarked about a year ago so i'll try to include them where applicable but there is of course a big asterisk there because those numbers were collected a while ago and unfortunately i don't have an m1 device that i can collect new numbers for so first let's start with geekbench the classic as you can see no surprise the m1 max macbook pro comes out on top the single core performance across all three devices is not all that different but you'll see the m1 max shine in the multi-core performance because of course it has more cores than the other two devices it's also interesting to look at the rosetta numbers because obviously the performance is going to take a hit when you're running under the rosetta 2 translation layer but the m1 max macbook pro is still much more powerful than my old 2018 15-inch next we'll talk about the compute test which runs on the gpu and of course there's no contest here the m1 max has an incredibly beefy gpu that's going to wipe the floor with the m1 and the intel macbook pro whether you're running on metal which is more optimized or opencl which is the more general standard now let's get into some benchmarks that programmers specifically may be interested in i downloaded the latest xcode beta and had both laptops unarchive the xzip file this is a cpu bound task and unarchiving can take advantage of multi-threading as well so you'll see that the m1 max macbook pro finished in less than half the time that the 2018 15-inch macbook pro took now let's take a look at some compilation tests on both machines i used the homebrew package manager and for each package i had it build from source instead of downloading the pre-built binary as you can see from the graph the m1 max is somewhere in the ballpark of twice as fast as the intel machine so sometimes it's a little more than twice and sometimes a bit less but there's definitely a noticeable performance improvement in the m1 max next i ran the pi performance benchmark this is the standard benchmark for python it runs lots of different python tasks from django to senpai symbolic math to regular expressions and i've picked a few of the results to include here as you can see the m1 max macbook pro is again a lot faster next i ran two javascript benchmarks jetstream 2 and kraken and i ran both on safari and chrome these javascript benchmarks are important not only if you write code in javascript but also if you visit websites on the internet interact with complex web apps or electron apps like spotify discord and slack no surprise but the m1 max gets the highest score in jetstream 2 and finishes the kraken benchmark the quickest it's also interesting to note that scores for safari are higher across the board compared to chrome both browsers are running natively but safari may be a bit better optimized for mac os and finally to let the m1 max truly flex i trained a machine learning model on both computers i went with the mnist digit classifier data set it contains 42 000 images of handwritten digits and i let it run on both machines the m1 max thanks to its neural engine finished way quicker at just an hour and 20 minutes compared to the intel mac which took 3 hours and 11 minutes the last thing i'll note is that while running a bunch of these benchmarks the intel macbook pro would consistently run the fans very loud and it would get pretty hot to the touch whereas the m1 max macbook pro didn't run the fans pretty much at all and it was generally cool or maybe warm to the touch so the thermals are looking a lot better [Music] now i want to talk a bit about the design and experience of the new macbook pros because as exciting as the performances there is an entire computer surrounding the new chip first i absolutely love the display going from a 15 inch to a 16 inch doesn't sound like a big difference but it is actually huge especially when the display can go up into the upper bezel plus of course having mini led and pro motion are huge pluses the display is just absolutely beautiful i'm also glad to see the touch bar gone it never really bothered me but i also never really liked it i only ever used it for brightness and volume adjustments and i can do those just fine with the standard function keys this new macbook pro is also noticeably thicker than my old macbook pro i think there was a trend back in 2016 and earlier where companies especially apple would try to release devices that were as thin as possible that brought us the iphone 6 and 6 plus and the whole ben gate scandal there and then of course the 2016 to 2018 macbook pros that had some pretty bad thermals due to how thin they were so i personally am totally fine with a thicker device if it means more power and proper thermals to support it i also say the better battery life is a huge benefit i usually use my laptop docked at a monitor but when i'm sitting on the couch or if i'm out somewhere it's really nice to have the much improved battery life especially when working on performance heavy tasks and then of course there are the additional ports less fan noise and instant wake from sleep and lots of other very nice quality of life features and so now with all of that in mind i'd like to give my final thoughts i want to answer two questions here first of all are the new 14-inch and 16-inch macbook pros worth upgrading to and second of all is it worth getting the fully specced m1 max versus the base model m1 pro or maybe something in the middle so to answer the first question yes these new macs are absolutely worth upgrading to you get the new design you get a bigger screen much better battery life and all of the awesome things that i mentioned before but you also get all of that stuff on the base model and i think for most people the base m1 pro will be more than enough power honestly the m1 from a year ago would probably be more than enough power take a second to think about what you actually use your computer for for me of course i use it for coding but i also use it for browsing the web checking my email watching videos and so on these are much lighter tasks that definitely don't need the power of the m1 max and even when i am writing code the majority of that time is spent in the ide actually typing code and that's not going to get any faster regardless of what processor you're using so it's only when i'm compiling and testing and running my code where i'd see a big improvement and will i see an improvement absolutely but will it make that big of a difference in the long run especially when you're talking about the m1 pro versus the m1 max i really don't think so so here's my final verdict if you already have an m1 mac you should only upgrade if you really want the new design or if your current computer can't handle your workload if you have an intel mac it's absolutely worth the upgrade but maybe not to the fully specced m1 max i would consider an m1 pro maybe even an upgraded m1 pro and that should be more than enough to handle the vast majority of programming related tasks so given my final verdict you may be wondering if i regret purchasing the fully specced m1 max macbook pro and the answer is kind of but i'm probably not going to return the computer i think i'd be perfectly fine with the base m1 pro i might upgrade it and i would also probably upgrade the unified memory and the storage space but i already bought this computer and having all this extra power means that it'll last me longer so i think i'm just gonna keep it if you have any questions related to programming on the new macbook pro feel free to leave them below if there's enough interest i may make a follow-up video and of course you can like the video if you liked it subscribe to my channel follow me on twitter you know the drill thanks so much for watching see you guys in the next video bye for now [Music] [Music] so given my final recommendation you may be wondering if i regret purchasing the fully maxed out m1 max macbook pro wow that was a tongue twister
Info
Channel: Noah Rubin
Views: 7,074
Rating: undefined out of 5
Keywords: m1, pro, max, m1x, Mac, book, MacBook, developer, code, coder, coding, Xcode, compile, compilation, test, benchmark, bench, mark, Geekbench, geek, create, ml, machine, learning, CreateML, python, pyperformance, performance, testing, air, 14, 16, inch, 14inch, 16inch, 2021, new, laptop, arm, x86, apple, silicon, jetstream, jetstream2, kraken, unzip
Id: AnBUHfbR_uU
Channel Id: undefined
Length: 13min 46sec (826 seconds)
Published: Sat Nov 06 2021
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.