First of all, thanks for putting Data Engineering in the list. Although we program a lot and share a lot of technologies with the devs, we are often forgotten in these tech videos. I work as a contractor for an American company and have to get myself the PCs I use for work. I like having a portable ultrabook in order to be able to work outside my home office, but I was frustrated with my Dell XPS 13 and decided to take the risk and get a MacBook Air M1. So far, most of my work is fine, since most of it is SaaS based, but when I had to run a couple of Docker containers with an older Apache Airflow instance, I ran into some troubles. The version they use is a bit older and even forcing it to build with an x86 instruction set, it didn't fully worked. Because of that, I am still keeping my Dell XPS for the times I need to run something locally, but that's only 5 to 10% of my work. What I like the most of my M1 machine is the battery life (I never saw something like that, it's amazing) and the fact it doesn't get hot unless I have to use Docker. Living in Brazil, a tropical country with high temperatures for most of the year, this is exactly what I was looking for. I hope that this comment can be useful to anyone in a similar position.
As a dev that also uses an m1 machine I can suggest that you can use a cheap VM in the cloud should you ever need to periodically run a docker container that is not compatible with ARM. Set up a AMD/x86_64 on AWS, GCP, or your favorite cloud provider then ssh in and build and run your container there. Just remember to shut the VM down when you don't need it!
@@MarvinTurner yeah probably use something like terraform to automate the infra part, and this method using cloud for specific things and just ssh with m1/m2 air and U can run basically anything
Me being a data engineer honestly never had performance problems from any laptops because if the laptop was only able to run: 1. Remote desktops 2. Able to run ssh shells 3. Able to run 10-15 browser tabs (for Snowflake and AWS) It was more than enough for me. Using a mid line Acer laptop of around $550 last 6 moths. (have used similarly specced laptops for last 7-8 yrs was more than enough for me.
Hi Alex, I wish I found your TH-cam channel earlier with this review. I purchased a MacBook Pro M1 Max 14" with 4 TB SSD, 32 GB RAM, and 32 GPU Cores. It was to replace my MacBook Pro i7 (2020). M1 Max does open every day stuff like e-mail, FaceTime and etc faster than my macbook pro i7. It does not however do anything for TH-cam or Amazon Prime Video. I am not a content creator, software developer, or video editor. I realized I have made a serious mistake in purchasing the M1 Max. It cost me USD 5800.00 for the notebook. For those reading this comment, please don't reply with negative comments to me as I already feel stupid enough to make such an expensive mistake. Keep up the good work Alex, I have already subscribed to your TH-cam channel.
As an Data Scientist (DS) i was a bit suprised by your choice. Because most DS don't use their own Notebook for creating ML Algorithm. In my Company we have a High End Server for that. And most of the time you don't expect the Results as soon as possible. Some Hours/Days are normal depending on the algorithm and the Data Set. So a M1/M2 MacBook Air is enough in my opinion.
As an Android developer, I had a chance to compare 13" Pro M1 and 14" Pro M1 Pro, both with 16Gb of RAM. M1 Pro is my personal machine and M1 is provided by the company I work for. When connected to an external monitor, keyboard and mouse, I didn’t notice real-life performance difference at all. Project that I’m working on builds from scratch 2m on M1 and 1m45s on M1 Pro. I’m okay with both, I guess :) The fact that the difference is small to nothing is good, because i am not allowed to work on personal laptop.
If you gonna get M1 pro no reason to go with less than 32gb for me really Those devices with everything is soldiered makes you really go max in stuff you can’t really work with M1 max is overkill for majority so you are left with M1 pro Ram can’t even expanded so max it out 32 And SSD you can have external drives
Does your 14" Pro M1 Pro come with 10 CPU cores, or 8 cores? I am trying to choose between 13" M1 8 cores and 14" M1 Pro 10 cores. Currently I have an experience with 13" M1 8 cores for Android development. Thanks.
I agree with the rankings for the most part, but I will say if you develop on Docker for Mac, my strong recommendation is to get a Pro chip with 32GB unified memory. I definitely go over 16GB on a regular basis as a full stack developer
I'm also need to buy a MB Pro for personal projects/online courses. I run several docker containers as Sidekiq, Redis, Postgres, app + api (eg Rails, Node, React). Will 16GB enough or 32 is better? Thanks!
@@tebe11 in my opinion, anyone using Docker for Mac should not get less than 32GB if at all possible. It’s probably the highest value upgrade you can get, arguably higher than storage but that may be up to your use case. Good luck!
As an iOS Developper, I moved from m1 to m1 pro, because m1 does not has enough ram management for bigger project (the 16 seems really short and always full), the IO does not handle enough things (screen + 1 external ssd seems to be my limit, I can not add an other ssd or screen without loosing one of them)
I plan to move from my old MacMini Intel i5 to a new M1 chip. I was waiting for the new M2 chip (but that did not convince me a lot). I am a web/backend developer. So after watching lots of comparison videos I came to the result that the M1 Pro with 16 GB RAM is the best fit for most of the requirements. I am following your comparison videos from the last 1 week and believe me your videos are much better than others discussing the M1 and M2. Thanks for your videos.
These days, CPU is almost never a limitation anymore. I've done all this work on a (maxed out) 2009 Mac Pro. Issue with the M1 is memory and video outputs. I run 2 or 3 monitors, depending on where I am, so the M1 is out. When I do back-end development (on different projects), I spin up my environments in Docker. To make things modular and re-usable, some projects require 5-8 containers at the same time (back-end, front-end, cache, database, workers, etc). Working with large-ish datasets will make you run out of the 16GB memory fast! Now if you've got multiple projects going on and/or need to debug communication between several (micro) services for example... Better be safe and spend the extra money on that memory, the hours and frustration you'll save not waiting and messing around are worth more. I'd say go with the M1 Pro at 32GB minimum, for any serious work. If you do anything 'corporate' (Oracle, .Net, etc), you basically have no choice, your employer will provide you an Intel based Windows machine.
agree 100%… i use a dual x5680 with 64gb ram 2010 Mac pro connected to 2 4k monitors, heavily using docker/kubernetes and still does the job fine… the sweet spot i think is a macbook pro m1 pro with AT LEAST 32gb of ram… especially for people doing full stack or both ios and android development using frameworks like react native, flutter, etc, using ide like intellij/android studio/xcode, will regret buying an m1 with max 16gb ram….
You can emulate pretty well on a MacBook Air with 8gigs, but yeah, I agree it can be kind of a pain, so if you can’t find the 16gb (which was my case) for a descent price (MacBooks are more expensive here than they are in the US, believe it or not) just get an iPhone SE 2 and you’ll be fine for emulating.
the best setup imo is a custom built PC + macbook air 16GB. the air is more than enough for pretty much most programming workflows, especially web development, and so splurging on an m1 pro machine is quite literally a waste of money because you’ll almost never use all that power. the PC would be great for gaming if you’re into that, where splurging on powerful components will actually be useful and noticeable. if you’re not into gaming and have money to spare i’d just go for a base 16” macbook pro
New subscriber here and backend software engineer here, thank you for your video! Ive been searching far and wide for this type of video! All these content creators have speced out M1 max's with 64 gigs of ram, 2TB SSD, and I honestly am jealous but I keep reminding myself.... I am not exporting 4k/5k/6k video... I love macs so much! but even the price tag on the M1 air is crazy (i paid it) but you get the same amount of performance with a slightly lesser price tag on a DELL XPS or a Ryzen computer.... But what you are paying for is not just the name with macs. You are paying for the top 3 things essentially... Battery, keyboard and trackpad... Let alone the price tag on the speced out M1 max? its like we are looking at used car price tag territory... $6,000 ridiculous... I paid $2,299 at best buy for my 2019 MBP intel i9 16gigs ram, and 1TB SSD back when it was on sale for $500 off in august 2021, and honestly its really not a bad machine for backend work. It handles my work flow for school fine, and handles my back end work stuff just fine... My company uses Java for their backend software so I have intelliJ, Postman and Studio3T for mongo DB stuff plus some google chrome tabs, slack application, discord app, music app, and i barley hit 12 or 13 gigs of ram... I do have yet to take my "capstone" classes at my university where we are doing a team full stack application, but i am almost positive my machine (2019 MBP intel i9 16gigs ram, and 1TB SSD) can handle that just fine. Moral of the story: I really want a M1Pro chip or even an M1 Max chip, but not willing to pay the price for it especially when my 2019 MBP intel i9 16gigs ram, and 1TB SSD can handle my school and work just fine I may treat myself come time august and bestbuy hopefully has them on sale for back to school for 100 dollars off... Then and only then I will treat myself and the model im looking at is a 16 inch M1 Pro chip with 10 core CPU and 16 core GPU 16 gigs ram and 1TB SSD, Thats perfect for my backend needs (I do not do video editing and 4k video exporting nor do i do photoshop or lightroom)
Cool review! One thing I would add is that for ML whilst it may be handy to have good embarked processing power, for bigger tasks I would rely on cloud services to scale up and down my processing power depending on the task at hand. This is much more economical and allows when needed for processing power much larger than anything that can be attached to any of the M1 chips.
Thanks for the great comparison Alex. I am looking into replacing my seemingly faulty XPS 15 with a Mac and your comparison is exactly what I need now. Sadly I also thought 16GB should be fine for Frontend development. Our environment runs with around ~14-16GB RAM usage on Windows. Now that I also do the backend code, having the 32GB machine really helps. Sure our use case is very specific but I'd like to be able to keep a machine longer than the current 2 years on my XPS, so for me having at least 32GB for "frontend" makes sense :)
Java doesn't work properly on the m1. Even if you use the azul or liberica jdk's that are native for the m1, there are still places where they need to call native code which hasn't been compiled for aarch64. If you're using the very latest jdk's etc, then maybe it's been fixed, but most projects are on java 8 jvm.
Excellent video. I bought M1 Air 8GB for primarily native iOS dev and occasional flutter dev. I upgrade from my old MBP 15" 2014 16GG model. I am power user meaning I use my laptop for everything. I also do Unity development for hybrid apps (not games, for games I have a high end windows PC) I was a 3D Modelling and texturing artist before I started my development career so I wanted to have that option too if required, incase I ever have to make something for my hobby (I have a high end PC for that if required) Now, believe me M1 is FAST and I really like the upgrade form my previous intel high end model. No fan noise and very cool in hand (occasionally it do gets warm for some time when using lot of programs) But 8GB is not worth it. DONT get the 8GB model if you are thinking long term. The only reason I got it is because I have Hackintosh too of 32Gb ram for backup and I wanted to see the power of M1 and have plan to upgrade it next year. Also I got this laptop in discount so its like a trial run. I OVE how light it is and portable (Have Ultrawide setup when I require screen real estate). It fits my need perfectly and I see my self getting rid of my Hackintosh with my next upgrade.
i am a full stack web developer and i have bought new M1 Air for quick/urgent bug fixes (if not available on my workstation) for clients and have Custom Xeon Station for all section at ones Graphics/UI/Backend/Database divided by dedicated 3x screens etc.
Nice job! I really liked the way you developed this comparison. Very unique. I've been doing full stack on the M1 Pro base model and it's fantastic. The only thing I am lacking is an Azure Cosmos DB Emulator (unsupported on ARM) but there are workarounds.
Thank you for a great information! But, as a game developer, I can say, that if you are making such a little games like hyper casual games or something - M1 2020 or Intel 2019 16G probably be the thing for you. But for bigger projects do not use m1 2020, use at least Intel 2019 mac book pro with 32 gigs of ram, or use m1 2021, but consider do not touch 16 gigs models - it's really small amount of memory even for middle-size casual projects.
In your last video you mentioned future proofing. I think this is a normal reaction for us devs. Make the computer last as long as it can, but for these M1 machines, I agree with The Every Dad when he said usable power is the name of the game.
Future-proofing is not always financially reasonable. it is often better to buy the "enough for now" and upgrade more often, especially if there is a person who could use an older machine - like a family member etc.
That’s a great high level overview! Im sure it fits a lot of developers. Another take is from d3vtec, who apparently works on Android apps, some of which are very large. He is now completely outfitted with Mac Studios, as well as M1Max MBPs, for building and testing the apps. You’d probably like it.
Full stack dev here. I use a base M1 air professionally. I bought an M1 Pro because it has a bigger, more beautiful screen and it can support my other monitors. If M1 Air could support more monitors, it would be a dream machine.
@@ko-Daegu Becuase it is, CPU-wise. I had a 2019 maxed out Intel MBP with 64GB Ram that died. Replaced with an 8GB Air. Didn’t miss the Intel in the slightest and was blown away at how the Air acted like it had all the Ram it needed. The key as covered by other channels is the fast SSDs which make swap space almost a mute point. Up until literally tonight with delivery of my M1 Pro 16GB 16” (I hear it has better thermals and performance), I have been able to do my full stack development - 2 Xcode projects (one being an app and the other a Swift Vapor backend) and a Docker container simultaneously on the Air 8GB with no issues… Docker compiles on my 10850K desktop aren’t significantly faster than the Air either… Also agree with OP I upgraded mainly for the bigger screen lol. The M1 Air is an absolute beast. I would have loved to see the M1 Pro in the Air but I guess it has beefy thermal requirements.
Please if you see this mate pursuing data science and currently doing full stack M3 PRO 18GB 11 CPU 14 512GB 14INCH OR M1 MAX 32GB 10CPU 32 GPU 1TB which should I go for ? both same price
Great breakdown and thanks for alowance to buy whatever I like as a mobile developer. I can even keep my gaming PV just in case I will build Android Apps :) Luckilly I'm not 10x engineer.
As a frontend dev I'd argue that it can get pretty heavy. For example, I'm using Figma (some might use the Adobe suite as well), running Docker in the background that's running a proxy, then VSCode, and often I also want to watch jest tests in the background, or even have E2E test suites run locally. To test and debug, I have different browsers and dev tools open in parallel, and sometimes I have to test in Android/iOS simulators. Luckily the need to test in Windows VMs with IE is a thing of the past. M1/M2/M3 non-Pro is something I wouldn't want to use, simply for the reason that I need 3 screens (code/design/result, or code/terminal with servers and tests/result). Until now, I'm using a 2019 i9 MBP, and it gets hot and loud quickly, especially with two external screens connected and Docker and/or jest watch running. I'll soon get an M3 Max, but I think that's overkill. That's just what my company always orders, so I don't complain. M3 Pro should be the sweet-spot now, and M1 Pro and M2 Pro should work just as well.
Best combo for me as a full stack engineer is any laptop and a remote build server. I dual boot my home gaming PC with Linux and when I am working I will use remote development tooling like the VSCode SSH extension on my laptop to essentially make my laptop computer appear to do the work locally, but it in fact offloading everything from intellisense to compilation to my beefy desktop PC. M1 Mac cannot output to more than one monitor so that's a bad thin client. The MacBook pro is better - a thin client just needs to be ergonomic to use, have good battery life and adequate performance. A bit pricey though, but probably the best choice at the moment.
As Mobile engineer working in iOS and Android I think that you need to work at least one day with a Mac - intel version and let us know what it is going to happen. On mobile I would suggest a totally different ranking M1 max first second M1 pro and the other the two left just forget about them. The reason is that we need to test a lot in different screen sizes and different OS version.
Hey Alex. You are the best! This is the perfect video for making a choice. I've been looking at a lot of TH-camrs covering the M1 Macbooks and this is the first video i've seen that has gone into detail about developer workflows and which macbook can tank them properly. You've convinced me to get an M1 Macbook Air (8+8) 16 GB RAM 512 GB SSD. I can't like this video enough.
@@anirbanc999 agreed. The 13" M1 MBP has been totally fine for me performance wise, running a large typescript monorepo. I yearn for a larger screen, but I actually returned the 16" due to weight/cost (I got the MAX though lol). The perfect laptop would be one that apple doesn't make. A 16" air that weighs around 4lbs.
But the 14” base model is way better value for money than the 16/512 m1 air is
2 ปีที่แล้ว
@@r3ndszergazda In my country (Poland) base 14" M1Pro 8c/14c 16GB/512GB costs 86% more than Air 8c/8c 16GB/512GB but I understand that the prices in US are different...
@ In my country (Hungary), base 14" M1 Pro costs 880k HUF and the 16/512 Air costs 610k HUF. That's 45% more, but then you get a way better screen (Air's screen is the same as the base iPod's screen & it is pretty outdated by now), 2 more performance cores and the ports. I think the base air is great value, but the 16/512 is not that great. But I guess it depends a lot on preferences & use cases, so it could be that you are right!
Good breakdown. FYI, probably the majority of the biggest “big data” stuff no longer runs on databases per se, but actually runs on top of Spark, which is a hefty JVM stack.
For android and in general why not go mba and a windows / linux desktop. You can always remote into the machine on mba and get amazing battery life and performance. You can get a decent desktop for $1k and an mba for around $1k. You’ll get way more performance than a 14 mbp. Intel macs are doa right now because they will have less software support.
For one, DevOps isn't person or position. This just - at core - really simple concept of organizing work of development and operations deparments. Joining two silos in one pipe.
@@AZisk Devops is for me a working concept in a team. The skill is distributed all over the team and should not rely on a specific person with this role. In my opinion, and its just my opinion. The industry is sometimes just rebranding the old fashioned "sys-admin" to devops. Because it sounds more cutting edge, if thats makes sense. Enjoyed the video good stuff!
Finaaaaalllyyyyyyyy Finallyyy!!!!!!!!!!!! One good video for software developers who have been working on either linux/windows(most probably linux) to see if there is any advantage with the m1 mac. I was fed up with aaaaaaaalllllllll the videos talking more about video editing,final cut pro and shit when the title mentions about software development. Thanks Alex. Thanks for keeping it to only SDE stuff and covering so many different developers!!!. Thank you so much.
As a full stack developer the best thing I ever did was get either a few monitors or one huge monitor and get the intel mac mini with 64gb of ram. This way you can read tons of docs open and switch between multiple environments.
Thanks for the video. I have a base model M1 MBA for web dev and it works. But I have been planning to upgrade and my first choice is the base model M1 Pro 16 inch
That mba is capable, but yes the newer m1 pro MacBooks are insanely powerful. Might be a bit overkill for web developers but they are future proof with better sound/camera/screen/keyboard. Worth it if you can afford it. If not, there’s that m1 mba and even the m1 mbp is great too.
@@corpuzone 6 months later I get work done but my workload has gone up beyond web apps. I'm now considering 32GB M1 Pro. Still looking. Still researching. I'll know after the new devices are announced later this year and reviews come in. But yes, base m1 air for web dev has really worked really, really well
I work as a data engineer. And vanilla m1 with 16 gigs of ram isn't enough for me. I mean it's ok, somewhat, but I constantly run out of memory with a bunch of tools and ide's open, most of those things are written in java and eat a lot of RAM. Also I am forced to run the intel JDK on the m1, instead of the native m1 jdk, because there are cases, for example writing an avro file, where a java library calls native C++ code, which hasn't been compiled for the m1, and I get an error, saying "[FAILED_TO_LOAD_NATIVE_LIBRARY] no native library is found for os.name=Mac and os.arch=aarch64", but. think it's been fixed in the later versions of the library that I use, but if I want to switch to it our entire infrastructure has to be moved to a newer version of Spark. So, in my opinion intel is the only good choice for a big data developer.
I think you missed the biggest differentiation between the three M1 variations: maximum RAM capacity. Many developer workflows are impacted by memory pressure. The main benefit of the M1 Max is you can go up to 64GB. The M1 Pro can only go up to 32GB, which is likely enough for most developers. The M1 tops out at 16GB, which is pretty constrained, especially if you are running Chrome, which is a memory hog. For most developer workloads, the memory bandwidth and additional GPU cores on an M1 Max won’t matter.
What about native desktop application developers? I could see some being OK with M1 and M1 Pro for Mac-only development, but I think Intel is still preferable if Windows development is required. The virtualization solutions are still lacking quite a bit. I actually decided to go M1 Pro and a separate Windows machine, but not as convenient as being able to do everything on a single laptop.
Good job but Front-end is not accurate, there are people who do design and front-end code, they deal with Illustrator , photoshop, sketch, Invision ... I'm still not sure M1 is enough or not but it's definitely heavier than a lot of those other applications you described
I moved an i7-8750h 24gb to an Air 8-Gpu 16gb for unity and swift. Unity render time takes twice less time on mac while gpu burning on intel laptop air doesn't care. So interesting.
I have a good deal of experience with data science and ML, and let me tell you: 99% of your work is going to be done on remote servers. Training models on your personal laptop is neither recommended, nor even possible, most of the time. So, all you need is really a very lightweight setup as most of the heavylifting will be done on remote servers that actually run the code.
Gonna throw this out there. If you're doing full-stack or backend development with any container framework (docker, docker-compose, etc), you really should consider getting 32GB of RAM, just to be comfortable running all those containers and your IDE, browsers, etc. You don't, for example, want to do many ElasticSearch experiments without a ton of memory.
So a 14 core gpu 16 go ram and 512 ssd will be more than enough for an entry level programmer and leave room for lots of growth later? Or would getting the 14 inch M1 Pro be overkill
To me as web dev, for m1 to be a viable option, docker must absolutely work properly on it (or have moby engine on Mac, or lxc to work like docker, or podman to support auto directories mapping)... And of course, I can't work with legecy codes like php56, older versions of node(e.g. sass works only with version 14, so 15 is going to throw some Python error at you)
So weird i'm watching this tonight because I just got my new mba m1 tonight. So ever since they released the MBP Pro 14 & 16 i've spent many days trying to decide what to get. Honestly i almost got a full specd unit not so much because i wanted an m1 max more because i wanted 64GB ram. So of course i kind of went backward here buying a mba. Honestly i had an old late 2013 13 rMBP that really need to be upgrade. I am a full stack engineer & devops. I work a LOT with docker and honestly docker just seems to be kind of a mess with apple silicon now. Also the VM story on docker is a mess: limited # of distros you can use, if you want to use windows you have to use WOA. So anyways i didn't want to spend 4000+ on a laptop that i couldn't fully utilize so i bought a mba as an interim device and hopefully in a couple of years everything is ironed out and i can look at ordering a full specd out unit again. So far the only thing i don't like is the key travel on the keyboard isn't as much as i would like it to be and its taking some getting use to.
Great vid! But you forgot to mention that the base M1 macs cannot natively connect to more than one external monitor. Which for front end devs like myself is a massive limitation.
The topic is valuable, so voted up. Yet, as for the GameDev part - it’s really lacking. Just one example: Desktop vs Mobile gamedev. For mobile all the “mobile dev” arguments apply, however for the Desktop games all the new M1s are too limited for now. On one hand, the M1 are really a breakthrough from the consumer’s point of view: much better performance, GPU, etc. But on the other hand, there is still no feasible way to develop AND debug a cross-platform game on a M1+ machine without Bootcamp+Windows support. At least as long as Apple doesn’t have anything near as useful as RenderDoc as a most prominent example.
Dear Alex, very interesting video. You didn’t mention Java, not sure if you don’t like it or you forgot about it. Like it or not Java is on top of another emergent technologies, and by nature multi-platform. The ability of having a SpringBoot app, and run a JAR regardless of CPU architecture is priceless. After +30 years of software development, I’d dare to say: Java is very strong with The Force.
This is so useful - thx you so much. As a informatics student it helps to understand some if not all roles i am going to choose from 😁👌🏻 And the best laptop 😂🎉
Will take this video with a grain of salt since there's now more support for Apple silicon. As a data engineer, I think the compute requirements must be at par with data scientists as they will sometimes be utilizing ML tools and containerization too. With that, extending to an M1 Pro or Max should be needed. An M1 might not be enough for those tasks.
Hey my friends, I’m an IT student so, I do some codes like machine learning tasks on my laptop. I have a little weird question, which one should I choose? 1. Macbook air m2 16gb 512gb 2. Macbook air m1 base model + iPad air5 m1 base model 3. (A little bit pricier combo): MacBook air m2 base model + iPad air5 m1 base model
It really depends with data engineers. There are some, who needs just web browser/terminal and runs everything remotely. But when you are like me - big data developer, then you need more than MBA. I really appreciate that I can run some of my pipelines locally for debugging. But it needs some CPU cores and also lots of RAM. So 16” intel MBP or M1 Pro with 32 gigs of RAM is must have. Will be soon switching to 14” M1 pro MBP so I hope that I will not find anything that would force me to return back to my loud intel 16” MBP. M1 max may be beneficial with it’s 64 gigs of RAM but the price difference is too high for only few occasions when I would need that much RAM.
Do you think you'll actually need that much ram though? Personally, with x86 I'd definitely go for 32gb for data science, but with ARM I'm thinking about just picking up the 16gb Pro model
with my limited experience, here's my 2 cents: if you need to emulate mobile app, get m1, if you need to compile / run program on your local machine, get intel(apple has no amd), if your laptop / desktop is terminal to remote server get m1.
Agreed on the DevOps only with M1. If needed you can spin VMs in the cloud, and most of the building will happen in the central CICD system of the team. Have been working for a Vite of a year now with M1 and, although I need to get my Intel one for some Docker buildings, that’s something that really has happened like twice
One of the "hidden" take aways I got. You explained exactly why Apple is still selling the Intel Mac mini, even 2 years in you still cannot do everything on Apple Silicon. I am the MDM admin who makes up my employers hardware specs. The reason we dont bother with M1 (and M2) is because upgrading the RAM makes the device a custom SKU which takes forever to come in, and limited to a single display natively. M1 Pro checks too many boxes for an off the shelf SKU.
Bottom line (if you do a bit of everything, or live in VS), get a decent intel laptop (win 11), like the new dell XPS15 which does everything very well except iOS/ Mac desktop. Get a Mac Mini to complement it and for fun (Air if need portability). Logitech MX keyboard and mouse to switch between by dragging mouse between screens If you are on the other side, and Visual Code based - then a MacBook Pro, and something like a XPS 13 (or a virtual online pc - cost adds up). VS for mac is supposedly native now as well?
Grab Parallels and support my channel: prf.hn/click/camref:1100libNI - Use coupon code: ZISKIND20
This is the Intel/M1/Pro/Max comparison video that we have been waiting for!!!!
Thanks for this Alex and thanks for the shoutout 🤙🏼🤙🏼
It might be worthwhile to do this same video occasionally when there is more support for Apple Silicon. Thanks man!
Not a bad idea
@@AZisk I would appreciate periodic updates
Agreed!
@@daniebello lá lá ¹¹⅓5 12
+1
First of all, thanks for putting Data Engineering in the list. Although we program a lot and share a lot of technologies with the devs, we are often forgotten in these tech videos.
I work as a contractor for an American company and have to get myself the PCs I use for work. I like having a portable ultrabook in order to be able to work outside my home office, but I was frustrated with my Dell XPS 13 and decided to take the risk and get a MacBook Air M1.
So far, most of my work is fine, since most of it is SaaS based, but when I had to run a couple of Docker containers with an older Apache Airflow instance, I ran into some troubles. The version they use is a bit older and even forcing it to build with an x86 instruction set, it didn't fully worked. Because of that, I am still keeping my Dell XPS for the times I need to run something locally, but that's only 5 to 10% of my work.
What I like the most of my M1 machine is the battery life (I never saw something like that, it's amazing) and the fact it doesn't get hot unless I have to use Docker. Living in Brazil, a tropical country with high temperatures for most of the year, this is exactly what I was looking for.
I hope that this comment can be useful to anyone in a similar position.
As a dev that also uses an m1 machine I can suggest that you can use a cheap VM in the cloud should you ever need to periodically run a docker container that is not compatible with ARM. Set up a AMD/x86_64 on AWS, GCP, or your favorite cloud provider then ssh in and build and run your container there. Just remember to shut the VM down when you don't need it!
@@MarvinTurner yeah probably use something like terraform to automate the infra part, and this method using cloud for specific things and just ssh with m1/m2 air and U can run basically anything
Than you man!
the thing is as data engineers we have to write lot of custom code if we are not allowed to use etl tools
Me being a data engineer honestly never had performance problems from any laptops because if the laptop was only able to run:
1. Remote desktops
2. Able to run ssh shells
3. Able to run 10-15 browser tabs (for Snowflake and AWS)
It was more than enough for me.
Using a mid line Acer laptop of around $550 last 6 moths. (have used similarly specced laptops for last 7-8 yrs was more than enough for me.
Hi Alex, I wish I found your TH-cam channel earlier with this review. I purchased a MacBook Pro M1 Max 14" with 4 TB SSD, 32 GB RAM, and 32 GPU Cores. It was to replace my MacBook Pro i7 (2020). M1 Max does open every day stuff like e-mail, FaceTime and etc faster than my macbook pro i7. It does not however do anything for TH-cam or Amazon Prime Video. I am not a content creator, software developer, or video editor. I realized I have made a serious mistake in purchasing the M1 Max. It cost me USD 5800.00 for the notebook. For those reading this comment, please don't reply with negative comments to me as I already feel stupid enough to make such an expensive mistake. Keep up the good work Alex, I have already subscribed to your TH-cam channel.
Lol...
As an Data Scientist (DS) i was a bit suprised by your choice. Because most DS don't use their own Notebook for creating ML Algorithm. In my Company we have a High End Server for that. And most of the time you don't expect the Results as soon as possible. Some Hours/Days are normal depending on the algorithm and the Data Set. So a M1/M2 MacBook Air is enough in my opinion.
As an Android developer, I had a chance to compare 13" Pro M1 and 14" Pro M1 Pro, both with 16Gb of RAM. M1 Pro is my personal machine and M1 is provided by the company I work for. When connected to an external monitor, keyboard and mouse, I didn’t notice real-life performance difference at all. Project that I’m working on builds from scratch 2m on M1 and 1m45s on M1 Pro. I’m okay with both, I guess :) The fact that the difference is small to nothing is good, because i am not allowed to work on personal laptop.
If you gonna get M1 pro no reason to go with less than 32gb for me really
Those devices with everything is soldiered makes you really go max in stuff you can’t really work with
M1 max is overkill for majority so you are left with M1 pro
Ram can’t even expanded so max it out 32
And SSD you can have external drives
@@ko-Daegu 16G is just fine, i develop for iOS and don't feel i need more, like it was when i had the touch bar mac.
Does your 14" Pro M1 Pro come with 10 CPU cores, or 8 cores?
I am trying to choose between 13" M1 8 cores and 14" M1 Pro 10 cores.
Currently I have an experience with 13" M1 8 cores for Android development.
Thanks.
@@pauldanchenko I have M1 Pro with 8 cores
@@pauldanchenkoWhich one did you get?
Finally, I have answer to “What’s a good laptop for a full stack developer”, currently studying for it. Thank you. I went with a MacBook Pro 15” 2015
I agree with the rankings for the most part, but I will say if you develop on Docker for Mac, my strong recommendation is to get a Pro chip with 32GB unified memory. I definitely go over 16GB on a regular basis as a full stack developer
Thanks my dude, this is what I was looking for
I concur with this, my docker setup takes 4GB alone. With everything else I’m running and the OS, 16GB would slow things down a lot. Go for 32GB.
Helpful comment, thank you
I'm also need to buy a MB Pro for personal projects/online courses. I run several docker containers as Sidekiq, Redis, Postgres, app + api (eg Rails, Node, React). Will 16GB enough or 32 is better? Thanks!
@@tebe11 in my opinion, anyone using Docker for Mac should not get less than 32GB if at all possible. It’s probably the highest value upgrade you can get, arguably higher than storage but that may be up to your use case. Good luck!
As an iOS Developper, I moved from m1 to m1 pro, because m1 does not has enough ram management for bigger project (the 16 seems really short and always full), the IO does not handle enough things (screen + 1 external ssd seems to be my limit, I can not add an other ssd or screen without loosing one of them)
this was very helpful ty! Did you get 8core or 10core? I'm deffs getting 32 ram cause I'm a ram hog
What configuration did you get?
By far the most detailed informative video I have ever seen. Post it to my college, since it answered all of my questoins.
Fantastic explaination.
I plan to move from my old MacMini Intel i5 to a new M1 chip. I was waiting for the new M2 chip (but that did not convince me a lot). I am a web/backend developer. So after watching lots of comparison videos I came to the result that the M1 Pro with 16 GB RAM is the best fit for most of the requirements.
I am following your comparison videos from the last 1 week and believe me your videos are much better than others discussing the M1 and M2.
Thanks for your videos.
I agree with you
What are you going to buy?
These days, CPU is almost never a limitation anymore. I've done all this work on a (maxed out) 2009 Mac Pro. Issue with the M1 is memory and video outputs.
I run 2 or 3 monitors, depending on where I am, so the M1 is out.
When I do back-end development (on different projects), I spin up my environments in Docker. To make things modular and re-usable, some projects require 5-8 containers at the same time (back-end, front-end, cache, database, workers, etc). Working with large-ish datasets will make you run out of the 16GB memory fast!
Now if you've got multiple projects going on and/or need to debug communication between several (micro) services for example... Better be safe and spend the extra money on that memory, the hours and frustration you'll save not waiting and messing around are worth more.
I'd say go with the M1 Pro at 32GB minimum, for any serious work. If you do anything 'corporate' (Oracle, .Net, etc), you basically have no choice, your employer will provide you an Intel based Windows machine.
agree 100%… i use a dual x5680 with 64gb ram 2010 Mac pro connected to 2 4k monitors, heavily using docker/kubernetes and still does the job fine… the sweet spot i think is a macbook pro m1 pro with AT LEAST 32gb of ram… especially for people doing full stack or both ios and android development using frameworks like react native, flutter, etc, using ide like intellij/android studio/xcode, will regret buying an m1 with max 16gb ram….
You can emulate pretty well on a MacBook Air with 8gigs, but yeah, I agree it can be kind of a pain, so if you can’t find the 16gb (which was my case) for a descent price (MacBooks are more expensive here than they are in the US, believe it or not) just get an iPhone SE 2 and you’ll be fine for emulating.
I watched practically the whole video and most if not all of it went over my head. But its a good video that I’m going to have to watch again.
Been trying to get ahold of you on WhatsApp but I’m not having any success.
Time to renew this now! With M2 and most of the software now supported
the best setup imo is a custom built PC + macbook air 16GB. the air is more than enough for pretty much most programming workflows, especially web development, and so splurging on an m1 pro machine is quite literally a waste of money because you’ll almost never use all that power.
the PC would be great for gaming if you’re into that, where splurging on powerful components will actually be useful and noticeable. if you’re not into gaming and have money to spare i’d just go for a base 16” macbook pro
New subscriber here and backend software engineer here, thank you for your video! Ive been searching far and wide for this type of video! All these content creators have speced out M1 max's with 64 gigs of ram, 2TB SSD, and I honestly am jealous but I keep reminding myself.... I am not exporting 4k/5k/6k video... I love macs so much! but even the price tag on the M1 air is crazy (i paid it) but you get the same amount of performance with a slightly lesser price tag on a DELL XPS or a Ryzen computer.... But what you are paying for is not just the name with macs. You are paying for the top 3 things essentially... Battery, keyboard and trackpad... Let alone the price tag on the speced out M1 max? its like we are looking at used car price tag territory... $6,000 ridiculous...
I paid $2,299 at best buy for my 2019 MBP intel i9 16gigs ram, and 1TB SSD back when it was on sale for $500 off in august 2021, and honestly its really not a bad machine for backend work. It handles my work flow for school fine, and handles my back end work stuff just fine... My company uses Java for their backend software so I have intelliJ, Postman and Studio3T for mongo DB stuff plus some google chrome tabs, slack application, discord app, music app, and i barley hit 12 or 13 gigs of ram...
I do have yet to take my "capstone" classes at my university where we are doing a team full stack application, but i am almost positive my machine (2019 MBP intel i9 16gigs ram, and 1TB SSD) can handle that just fine.
Moral of the story: I really want a M1Pro chip or even an M1 Max chip, but not willing to pay the price for it especially when my 2019 MBP intel i9 16gigs ram, and 1TB SSD can handle my school and work just fine
I may treat myself come time august and bestbuy hopefully has them on sale for back to school for 100 dollars off... Then and only then I will treat myself and the model im looking at is a 16 inch M1 Pro chip with 10 core CPU and 16 core GPU 16 gigs ram and 1TB SSD, Thats perfect for my backend needs (I do not do video editing and 4k video exporting nor do i do photoshop or lightroom)
Cool review! One thing I would add is that for ML whilst it may be handy to have good embarked processing power, for bigger tasks I would rely on cloud services to scale up and down my processing power depending on the task at hand. This is much more economical and allows when needed for processing power much larger than anything that can be attached to any of the M1 chips.
Same for the comments on Docker and even for building on mobile as well.
Thanks for the great comparison Alex. I am looking into replacing my seemingly faulty XPS 15 with a Mac and your comparison is exactly what I need now.
Sadly I also thought 16GB should be fine for Frontend development. Our environment runs with around ~14-16GB RAM usage on Windows. Now that I also do the backend code, having the 32GB machine really helps. Sure our use case is very specific but I'd like to be able to keep a machine longer than the current 2 years on my XPS, so for me having at least 32GB for "frontend" makes sense :)
This was the most helpful video iv ever seen about this very specific topic, thank u man !
As a backend developer, I have an M1 Mac, and M1 pro Mac, and M1 pro builds giant projects way faster than M1.
Do you have any issues deploying docker images?
Full stack developer here, using a base m1 MacBook Air. The only time this machine struggles is if I start the android simulator.
2017 mabook pro 15; yes, i am sadly when every time start android simulator
Front-end Dev here, I'm using MacBook Pro from 2018' with 16GB and Intel i5 (4x2,2GHz). I have this laptop since 2019 and it's good enought :)
As a Data Engineer, I think M1 pro would also be a good choice. Data Processing and ETL using Spark or Hive needs a good CPU performance imo
Java doesn't work properly on the m1. Even if you use the azul or liberica jdk's that are native for the m1, there are still places where they need to call native code which hasn't been compiled for aarch64. If you're using the very latest jdk's etc, then maybe it's been fixed, but most projects are on java 8 jvm.
I would think it's IO-bound not CPU-bound
Excellent video. I bought M1 Air 8GB for primarily native iOS dev and occasional flutter dev. I upgrade from my old MBP 15" 2014 16GG model. I am power user meaning I use my laptop for everything. I also do Unity development for hybrid apps (not games, for games I have a high end windows PC) I was a 3D Modelling and texturing artist before I started my development career so I wanted to have that option too if required, incase I ever have to make something for my hobby (I have a high end PC for that if required)
Now, believe me M1 is FAST and I really like the upgrade form my previous intel high end model. No fan noise and very cool in hand (occasionally it do gets warm for some time when using lot of programs) But 8GB is not worth it. DONT get the 8GB model if you are thinking long term. The only reason I got it is because I have Hackintosh too of 32Gb ram for backup and I wanted to see the power of M1 and have plan to upgrade it next year. Also I got this laptop in discount so its like a trial run. I OVE how light it is and portable (Have Ultrawide setup when I require screen real estate). It fits my need perfectly and I see my self getting rid of my Hackintosh with my next upgrade.
Thanks for sharing! Really fun to hear as I've never played with a Hackintosh and previously had a M1 Air with 8Gb
Love how you have to put a trigger warning for disagreements lol. Great info Alex!
i am a full stack web developer and i have bought new M1 Air for quick/urgent bug fixes (if not available on my workstation) for clients and have Custom Xeon Station for all section at ones Graphics/UI/Backend/Database divided by dedicated 3x screens etc.
Nice job! I really liked the way you developed this comparison. Very unique. I've been doing full stack on the M1 Pro base model and it's fantastic. The only thing I am lacking is an Azure Cosmos DB Emulator (unsupported on ARM) but there are workarounds.
Hi Brady, are you using the m1 Pro 8 or 10 core? thanks :)
Cassandra is an option. Not the same type of No-SQL, but still good.
Thank you for a great information!
But, as a game developer, I can say, that if you are making such a little games like hyper casual games or something - M1 2020 or Intel 2019 16G probably be the thing for you. But for bigger projects do not use m1 2020, use at least Intel 2019 mac book pro with 32 gigs of ram, or use m1 2021, but consider do not touch 16 gigs models - it's really small amount of memory even for middle-size casual projects.
This video is what I needed. Thanks a mil!
In your last video you mentioned future proofing. I think this is a normal reaction for us devs. Make the computer last as long as it can, but for these M1 machines, I agree with
The Every Dad when he said usable power is the name of the game.
Future-proofing is not always financially reasonable. it is often better to buy the "enough for now" and upgrade more often, especially if there is a person who could use an older machine - like a family member etc.
That’s a great high level overview! Im sure it fits a lot of developers. Another take is from d3vtec, who apparently works on Android apps, some of which are very large. He is now completely outfitted with Mac Studios, as well as M1Max MBPs, for building and testing the apps. You’d probably like it.
Full stack dev here. I use a base M1 air professionally. I bought an M1 Pro because it has a bigger, more beautiful screen and it can support my other monitors.
If M1 Air could support more monitors, it would be a dream machine.
Well You can with an external dongle
What do you do exactly??
How is the M1 air sufficient enough for a full stack work
@@ko-Daegu Becuase it is, CPU-wise. I had a 2019 maxed out Intel MBP with 64GB Ram that died. Replaced with an 8GB Air. Didn’t miss the Intel in the slightest and was blown away at how the Air acted like it had all the Ram it needed. The key as covered by other channels is the fast SSDs which make swap space almost a mute point. Up until literally tonight with delivery of my M1 Pro 16GB 16” (I hear it has better thermals and performance), I have been able to do my full stack development - 2 Xcode projects (one being an app and the other a Swift Vapor backend) and a Docker container simultaneously on the Air 8GB with no issues… Docker compiles on my 10850K desktop aren’t significantly faster than the Air either…
Also agree with OP I upgraded mainly for the bigger screen lol. The M1 Air is an absolute beast. I would have loved to see the M1 Pro in the Air but I guess it has beefy thermal requirements.
Very nice comparison I'm a backend developer and I agree with your suggestion 👍
favourite TH-camr now hands down
Please if you see this mate pursuing data science and currently doing full stack M3 PRO 18GB 11 CPU 14 512GB 14INCH OR M1 MAX 32GB 10CPU 32 GPU 1TB which should I go for ? both same price
you are my fav TH-camr 🙂
Great breakdown and thanks for alowance to buy whatever I like as a mobile developer. I can even keep my gaming PV just in case I will build Android Apps :)
Luckilly I'm not 10x engineer.
As a frontend dev I'd argue that it can get pretty heavy. For example, I'm using Figma (some might use the Adobe suite as well), running Docker in the background that's running a proxy, then VSCode, and often I also want to watch jest tests in the background, or even have E2E test suites run locally. To test and debug, I have different browsers and dev tools open in parallel, and sometimes I have to test in Android/iOS simulators. Luckily the need to test in Windows VMs with IE is a thing of the past.
M1/M2/M3 non-Pro is something I wouldn't want to use, simply for the reason that I need 3 screens (code/design/result, or code/terminal with servers and tests/result). Until now, I'm using a 2019 i9 MBP, and it gets hot and loud quickly, especially with two external screens connected and Docker and/or jest watch running. I'll soon get an M3 Max, but I think that's overkill. That's just what my company always orders, so I don't complain. M3 Pro should be the sweet-spot now, and M1 Pro and M2 Pro should work just as well.
Best combo for me as a full stack engineer is any laptop and a remote build server.
I dual boot my home gaming PC with Linux and when I am working I will use remote development tooling like the VSCode SSH extension on my laptop to essentially make my laptop computer appear to do the work locally, but it in fact offloading everything from intellisense to compilation to my beefy desktop PC.
M1 Mac cannot output to more than one monitor so that's a bad thin client. The MacBook pro is better - a thin client just needs to be ergonomic to use, have good battery life and adequate performance. A bit pricey though, but probably the best choice at the moment.
As Mobile engineer working in iOS and Android I think that you need to work at least one day with a Mac - intel version and let us know what it is going to happen. On mobile I would suggest a totally different ranking M1 max first second M1 pro and the other the two left just forget about them. The reason is that we need to test a lot in different screen sizes and different OS version.
Im a full stack engineer and use my M1 Air 16gb for everything.
How big your sdd?
1tb
Hey Alex. You are the best! This is the perfect video for making a choice. I've been looking at a lot of TH-camrs covering the M1 Macbooks and this is the first video i've seen that has gone into detail about developer workflows and which macbook can tank them properly. You've convinced me to get an M1 Macbook Air (8+8) 16 GB RAM 512 GB SSD. I can't like this video enough.
What about the screen size.. 13” is too small. 14” M1 MacBook Air would be perfect but we don’t have that sadly :(
@@anirbanc999 agreed. The 13" M1 MBP has been totally fine for me performance wise, running a large typescript monorepo. I yearn for a larger screen, but I actually returned the 16" due to weight/cost (I got the MAX though lol).
The perfect laptop would be one that apple doesn't make. A 16" air that weighs around 4lbs.
But the 14” base model is way better value for money than the 16/512 m1 air is
@@r3ndszergazda In my country (Poland) base 14" M1Pro 8c/14c 16GB/512GB costs 86% more than Air 8c/8c 16GB/512GB but I understand that the prices in US are different...
@ In my country (Hungary), base 14" M1 Pro costs 880k HUF and the 16/512 Air costs 610k HUF. That's 45% more, but then you get a way better screen (Air's screen is the same as the base iPod's screen & it is pretty outdated by now), 2 more performance cores and the ports. I think the base air is great value, but the 16/512 is not that great. But I guess it depends a lot on preferences & use cases, so it could be that you are right!
Spend more time watching these videos than coding, even though I was already convinced to get the m1
Thanks ; this was a really thoughtful video to share . Going to help a lot of people
One pro for the m1 pro, is support for two 5k 27" monitors. (Now that Apple released their displays)
Ah!!! Thanks... been looking for a video like this!!
Good breakdown. FYI, probably the majority of the biggest “big data” stuff no longer runs on databases per se, but actually runs on top of Spark, which is a hefty JVM stack.
Very helpful man, thanks
I am using air m1 16gb for mobile dev and it's more than enough :D, Thanks for video
Cool, does it work well for emulator on XCode and Android studio ? Planning to buy M1 Air 16GB for full stack + mobile dev
@@sureshpaulrajan Also compatible android emulator for m1 chip is released and it is work smoothly
For android and in general why not go mba and a windows / linux desktop. You can always remote into the machine on mba and get amazing battery life and performance. You can get a decent desktop for $1k and an mba for around $1k. You’ll get way more performance than a 14 mbp. Intel macs are doa right now because they will have less software support.
I love this guy!!!
For one, DevOps isn't person or position. This just - at core - really simple concept of organizing work of development and operations deparments. Joining two silos in one pipe.
Why you gotta go and insult all those DevOps engineers? They are people too! 🤪
@@AZisk Devops is for me a working concept in a team. The skill is distributed all over the team and should not rely on a specific person with this role. In my opinion, and its just my opinion. The industry is sometimes just rebranding the old fashioned "sys-admin" to devops. Because it sounds more cutting edge, if thats makes sense. Enjoyed the video good stuff!
Finaaaaalllyyyyyyyy Finallyyy!!!!!!!!!!!! One good video for software developers who have been working on either linux/windows(most probably linux) to see if there is any advantage with the m1 mac. I was fed up with aaaaaaaalllllllll the videos talking more about video editing,final cut pro and shit when the title mentions about software development. Thanks Alex. Thanks for keeping it to only SDE stuff and covering so many different developers!!!. Thank you so much.
As a full stack developer the best thing I ever did was get either a few monitors or one huge monitor and get the intel mac mini with 64gb of ram. This way you can read tons of docs open and switch between multiple environments.
Thanks for the video. I have a base model M1 MBA for web dev and it works. But I have been planning to upgrade and my first choice is the base model M1 Pro 16 inch
That mba is capable, but yes the newer m1 pro MacBooks are insanely powerful. Might be a bit overkill for web developers but they are future proof with better sound/camera/screen/keyboard. Worth it if you can afford it. If not, there’s that m1 mba and even the m1 mbp is great too.
@@corpuzone 6 months later I get work done but my workload has gone up beyond web apps. I'm now considering 32GB M1 Pro. Still looking. Still researching. I'll know after the new devices are announced later this year and reviews come in. But yes, base m1 air for web dev has really worked really, really well
I dont fully agree as I use a MacBook M1 Pro with 32Gb RAM for daily devops tasks and it works 100% fine for me
I work as a data engineer. And vanilla m1 with 16 gigs of ram isn't enough for me. I mean it's ok, somewhat, but I constantly run out of memory with a bunch of tools and ide's open, most of those things are written in java and eat a lot of RAM. Also I am forced to run the intel JDK on the m1, instead of the native m1 jdk, because there are cases, for example writing an avro file, where a java library calls native C++ code, which hasn't been compiled for the m1, and I get an error, saying "[FAILED_TO_LOAD_NATIVE_LIBRARY] no native library is found for os.name=Mac and os.arch=aarch64", but. think it's been fixed in the later versions of the library that I use, but if I want to switch to it our entire infrastructure has to be moved to a newer version of Spark. So, in my opinion intel is the only good choice for a big data developer.
Not many people are going to need more than a base m1 tbh.
I think you missed the biggest differentiation between the three M1 variations: maximum RAM capacity.
Many developer workflows are impacted by memory pressure.
The main benefit of the M1 Max is you can go up to 64GB.
The M1 Pro can only go up to 32GB, which is likely enough for most developers.
The M1 tops out at 16GB, which is pretty constrained, especially if you are running Chrome, which is a memory hog.
For most developer workloads, the memory bandwidth and additional GPU cores on an M1 Max won’t matter.
What about native desktop application developers?
I could see some being OK with M1 and M1 Pro for Mac-only development, but I think Intel is still preferable if Windows development is required. The virtualization solutions are still lacking quite a bit. I actually decided to go M1 Pro and a separate Windows machine, but not as convenient as being able to do everything on a single laptop.
Amazing video. Thanks for the insight.
Amazing Video!! Thanks for it..
Good job but Front-end is not accurate, there are people who do design and front-end code, they deal with Illustrator , photoshop, sketch, Invision ... I'm still not sure M1 is enough or not but it's definitely heavier than a lot of those other applications you described
I think for SQL Server, there’s Azure SQL Edge that you can use in Docker and M1s
I moved an i7-8750h 24gb to an Air 8-Gpu 16gb for unity and swift. Unity render time takes twice less time on mac while gpu burning on intel laptop air doesn't care. So interesting.
My 2020 M1 8GB slows down dramatically when node is running... So I would suggest either M1 with at least 16GB or one of the higher brothers...
I have a good deal of experience with data science and ML, and let me tell you: 99% of your work is going to be done on remote servers. Training models on your personal laptop is neither recommended, nor even possible, most of the time. So, all you need is really a very lightweight setup as most of the heavylifting will be done on remote servers that actually run the code.
Learning alot from you Alex, heres to hoping that you’ll do an update video on this - now that we have more support for the apple silicon chips
Yes Alex I am a developer and i think the minimum laptop should be the M1 pro because it has ports .😅 case closed
My final thoughts hahahahaha
Great video, thanks
Gonna throw this out there.
If you're doing full-stack or backend development with any container framework (docker, docker-compose, etc), you really should consider getting 32GB of RAM, just to be comfortable running all those containers and your IDE, browsers, etc. You don't, for example, want to do many ElasticSearch experiments without a ton of memory.
Yeah, 32 is a more comfortable amount, for sure.
Thank you! Look forward for updates when you have the Mac Studio setup too.
Awesome video like always! This comment is also for the algorithm. Keep up your good work.
This is some great content
So a 14 core gpu 16 go ram and 512 ssd will be more than enough for an entry level programmer and leave room for lots of growth later? Or would getting the 14 inch M1 Pro be overkill
To me as web dev, for m1 to be a viable option, docker must absolutely work properly on it (or have moby engine on Mac, or lxc to work like docker, or podman to support auto directories mapping)... And of course, I can't work with legecy codes like php56, older versions of node(e.g. sass works only with version 14, so 15 is going to throw some Python error at you)
Cute pfp
have you switched to m1 / m2? i'm also interested in to get proper battery life & performance developing on php7+, docker, older npm, etc
So weird i'm watching this tonight because I just got my new mba m1 tonight. So ever since they released the MBP Pro 14 & 16 i've spent many days trying to decide what to get. Honestly i almost got a full specd unit not so much because i wanted an m1 max more because i wanted 64GB ram. So of course i kind of went backward here buying a mba. Honestly i had an old late 2013 13 rMBP that really need to be upgrade. I am a full stack engineer & devops. I work a LOT with docker and honestly docker just seems to be kind of a mess with apple silicon now. Also the VM story on docker is a mess: limited # of distros you can use, if you want to use windows you have to use WOA. So anyways i didn't want to spend 4000+ on a laptop that i couldn't fully utilize so i bought a mba as an interim device and hopefully in a couple of years everything is ironed out and i can look at ordering a full specd out unit again. So far the only thing i don't like is the key travel on the keyboard isn't as much as i would like it to be and its taking some getting use to.
can you please update this video once m2 comes out if new support has been added for Apple Silicon. Really good video. Thanks
Nice Video
Thanks for making this ☺❤🙏
My pleasure 😊
Great vid! But you forgot to mention that the base M1 macs cannot natively connect to more than one external monitor. Which for front end devs like myself is a massive limitation.
They can if you use the right USB device... That's been covered plenty of times by now.
@@dankosek2355 I didn't know that, tbh.
The topic is valuable, so voted up. Yet, as for the GameDev part - it’s really lacking. Just one example: Desktop vs Mobile gamedev. For mobile all the “mobile dev” arguments apply, however for the Desktop games all the new M1s are too limited for now. On one hand, the M1 are really a breakthrough from the consumer’s point of view: much better performance, GPU, etc. But on the other hand, there is still no feasible way to develop AND debug a cross-platform game on a M1+ machine without Bootcamp+Windows support. At least as long as Apple doesn’t have anything near as useful as RenderDoc as a most prominent example.
Yes, these are good points
Hi Alex, always success for you
Thanks 🙏
I’m looking forward to seeing the next episode of these series! Apple has already released M2 Macs!
Dear Alex, very interesting video. You didn’t mention Java, not sure if you don’t like it or you forgot about it. Like it or not Java is on top of another emergent technologies, and by nature multi-platform. The ability of having a SpringBoot app, and run a JAR regardless of CPU architecture is priceless. After +30 years of software development, I’d dare to say: Java is very strong with The Force.
I have just started learning Java
Will the ARM be a problem ? Will i be forced to use parallels windows apps ?
@@nicknickname353 I use java for mobile development with android and it works just fine
This is so useful - thx you so much. As a informatics student it helps to understand some if not all roles i am going to choose from 😁👌🏻
And the best laptop 😂🎉
Will take this video with a grain of salt since there's now more support for Apple silicon. As a data engineer, I think the compute requirements must be at par with data scientists as they will sometimes be utilizing ML tools and containerization too. With that, extending to an M1 Pro or Max should be needed. An M1 might not be enough for those tasks.
We Need an updated videooo nowwww...
Hey my friends, I’m an IT student so, I do some codes like machine learning tasks on my laptop.
I have a little weird question, which one should I choose?
1. Macbook air m2 16gb 512gb
2. Macbook air m1 base model
+
iPad air5 m1 base model
3. (A little bit pricier combo):
MacBook air m2 base model
+
iPad air5 m1 base model
It really depends with data engineers. There are some, who needs just web browser/terminal and runs everything remotely. But when you are like me - big data developer, then you need more than MBA. I really appreciate that I can run some of my pipelines locally for debugging. But it needs some CPU cores and also lots of RAM. So 16” intel MBP or M1 Pro with 32 gigs of RAM is must have.
Will be soon switching to 14” M1 pro MBP so I hope that I will not find anything that would force me to return back to my loud intel 16” MBP. M1 max may be beneficial with it’s 64 gigs of RAM but the price difference is too high for only few occasions when I would need that much RAM.
Do you think you'll actually need that much ram though? Personally, with x86 I'd definitely go for 32gb for data science, but with ARM I'm thinking about just picking up the 16gb Pro model
with my limited experience, here's my 2 cents: if you need to emulate mobile app, get m1, if you need to compile / run program on your local machine, get intel(apple has no amd), if your laptop / desktop is terminal to remote server get m1.
I'm a front-end developer but I also run Visual Studio for my backend server environment. Does any of M1 macs will be able to handle this stack?
Why not just use the cloud for a heavy workload? How can you expect a PC to do an ML workload?
Agreed on the DevOps only with M1. If needed you can spin VMs in the cloud, and most of the building will happen in the central CICD system of the team. Have been working for a Vite of a year now with M1 and, although I need to get my Intel one for some Docker buildings, that’s something that really has happened like twice
One of the "hidden" take aways I got. You explained exactly why Apple is still selling the Intel Mac mini, even 2 years in you still cannot do everything on Apple Silicon. I am the MDM admin who makes up my employers hardware specs. The reason we dont bother with M1 (and M2) is because upgrading the RAM makes the device a custom SKU which takes forever to come in, and limited to a single display natively. M1 Pro checks too many boxes for an off the shelf SKU.
Thanks 🙏 good video
Docker can and probably used for every type of developer. Especially with VS Code dev containers.
Bottom line (if you do a bit of everything, or live in VS), get a decent intel laptop (win 11), like the new dell XPS15 which does everything very well except iOS/ Mac desktop.
Get a Mac Mini to complement it and for fun (Air if need portability). Logitech MX keyboard and mouse to switch between by dragging mouse between screens
If you are on the other side, and Visual Code based - then a MacBook Pro, and something like a XPS 13 (or a virtual online pc - cost adds up).
VS for mac is supposedly native now as well?
MacBook Air M1 16GB RAM is it enough for Flutter development?
should be good
Pro for professionals, Max for Maximalists 😊