Is MATLAB supported on Apple Silicon Macs?

Macs with an Apple silicon processor such as the Apple M1 have two ways to run applications. Applications built for Apple silicon run natively, while apps built for Intel processors run in the Rosetta 2 environment. Which method does MATLAB support? 

Answer 1

[MathWorks Support Team]

Rosetta 2 

MATLAB R2020b Update 3 and later are supported on Apple silicon Macs through the Rosetta 2 translation environment. Earlier releases of MATLAB may run but are not fully supported. 

There are two products which are available on Intel-based Macs but are not compatible with the Rosetta 2 environment: Polyspace and Simulink Desktop Real-Time. 

[Xiao Zang]

Kudos to Matlab team for the quick move!!

[Ben Joyce]

Yes, this is great news.

[Zonghao zou]

Do we have any ideas as in when the native apple silicon version will be avaliable? (sometime next year? )

[cr]

@MathWorksSupportTeam

When will Matlab for native M1 be available? Sharing some kind of outlook would be immensely helpful in making better decisions.

[Michel Georges Decker]

Probably a dumb question but: is Simulink Desktop Real-Time needed for Models that are not calculated in real-time?

[cr]

No, it's not required.

[Mohammad Sami]

For the optimized versions, would MATLAB be use the built in neural engine and machine learning accelerator for the deep learning toolbox and statistics and machine learning toolbox?

[Irene Mendez Guerra]

Will the optimized versions enable GPU acceleration on Apple silicon?

[Ravi Kohli]

Does this mean that MATLAB doesn't work on the new apple silicon devices for the time being?

[John Johnson]

"Kudos to Matlab team for the quick move!!"

Apple Silicon has been available to developers since last summer.

[Walter Roberson]

Developer Transition Kit was announced June 22, 2020, which was two days into Summer. Some developers received it about June 28, 2020 or June 29, 2020 https://9to5mac.com/2020/06/29/first-benchmarks-surface-for-apples-arm-based-developer-transition-kit/

So if "last summer" refers to the summer most immediately past, then Yes, it was available to developers "since last summer".

The R2020b Pre-Release was available to MATLAB beta test users no later than June 10, 2020.... which was before Apple Silicon became available. Mathworks policy is that after a pre-release has been made available to customers, that bug fixes may be made, and new features under trial may be removed before actual release, but that no new features may go in: internal Mathworks developers who do not have their features ready for external testing by the internal cut-off date have to wait for the next release (or at least an Update), even if the feature becomes internally ready even just a few days later.

The official R2020b release date was September 16, 2020. The official release date of the Apple M1 to consumers was September 17, 2020.

Thus, there was no realistic chance that Mathworks would have added Apple Silicon support to the R2020b release.

[Mike Bryant]

When will MATLAB R2020b Update 3 be available for download? I just tried to download and the latest available is Update 2.

[Bruno Luong]

Release 3 is already available since few days (for Windows). You might need you check with Help->Check for Update

[Winneyy]

How can I get in touch with up-to-date information or updates concering the availability of native MATLAB for MacBook Silicon?, through this post or through official post on your website?

[Luke]

Please do update this answer with more information as it becomes available. In particular: will 2021b have ARM support?

[Walter Roberson]

R2020b already has ARM support -- through Rosetta2 .

[BISHAL Kr DAS]

when can I expect the Native Silicon version to come out?

[Rik]

I doubt you will get anything more than speculation. Mathworks will probably mention it in the release notes for the prerelease (note that in general that would be under NDA).

[LeChat]

Will Matlab next version enable GPU acceleration on Apple silicon? Will the Deep Learning and Image processing toolboxes take advantage of the M1 integrated GPUs and neural engine?

[Rik]

@LeChat Mathworks generally doesn't discuss plans for the future in public (you could contact support, who will either refuse to talk specifics, or offer an NDA). Some next version of Matlab will probably enable GPU acceleration on Apple ARM chips, but I doubt R2021b will be that version. As R2021a is only working through Rosetta 2, I expect full support of more niche features (i.e. everything except the base product) will take a while.

[Walter Roberson]

"Will Matlab next version enable GPU acceleration on Apple silicon?"

No. Do not expect that for numerous years. Mathworks only supports Nvidia CUDA, and would have to completely redesign its gpu acceleration tool chain for anything else.

Mathworks goes where the market is, and research work is currently mostly on Nvidia, with second place in the market being IBM, and AMD based development far behind those two. Apple M1 development is comparatively small at the moment, and has no support for Windows or Linux. When Mathworks finally diversifies GPU it would make more sense for it to go after IBM (researchers) or AMD (commonly installed on desktops and laptops) before going after M1.

[LeChat]

Thank you both for your answers and views on Mathworks perspectives. For the mac users (actually with Intel cpu and latest macOS, or M1+ Silicon chips), do you know if it is possible to use a eGPU just for Matlab? I know Nvidia gpu drivers are not available in the latest MacOS versions anymore, and that nvidia gpus are not Metal compatible, but is there a built-in matlab support for a Nvidia eGPU option, so that the OS uses the integrated gpu of the computer while Matlab uses the eGPU acceleration?

[Walter Roberson]

There is no built-in support for egpu. MATLAB just uses whatever drivers are available

[Walter Roberson]

Also, the M1 requires Big Sur (or later), but there is no NVIDIA eGPU support for Big Sur, as Apple made it difficult to add shared drivers other than those approved by Apple -- and Apple apparently won't approve any eGPU drivers that are not designed for Metal.

[Ramy Elkhazendar]

Can we get this moving faster. Its been 5 months now! need a Native version ASAP! Performance gains are too good to pass

[Walter Roberson]

Well, 7 months if you count from the time that the M1 became available to consumers. One release has happened, and it improved the M1 performance and made more of it native. There will not be another release until close to the September equinox.

You should have a look, by the way, at Microsoft's efforts to implement for ARM. Their 64 bit OS for ARM barely works at present, and has not reached "beta" level.

[Justin Kottinger]

I am having major issues with MATLAB 2021a on my m1 Mac. The issues include: sudden crashes, unresponsive application, incorrect app loading, etc. The behavior is so terrible that I some days I give up trying to use the app entirely. Is anyone else having these issues? Will a native version be available anytime soon?

[Antony Ware]

I've started seeing crashes too, seemingly related to the editor. Selecting and pasting text often (almost always) results in an immediate crash. I'm running the latest 2021a on an M1 Macbook Pro. Anyone else?

[Joris]

Same issue here. Sudden crashes, seem to happen after pasting text.

Slow response to clicking with cursor, running and running sections. Where on my pc runnnig a section would be instant now it takes seconds. Especially the slow cursors is annoying, make selecting sections and text nearly impossible. Clicking on save now takes multiple seconds instead of instant response time.

Specs:

Macbook Air M1 16GB

Matlab R2021a

Any updates on M1 release?

[Anders Eriksson]

I can confirm that this issue has been appearing in R2020b as well. So it seems like there is something thats related to changes made by apple in macOS 11.3-11.4

[Dror Liran]

edit: I have updeted to update 5 and everything is much better

I experince the same issues with R2020b, is extreamly slow response.

Any solution for that?

MBP M1 16gb, macos 11.3

[Sangjae]

When will native M1 support be available? The current version is so slow that it can't be used in practice and often gets stuck. I'm currently switching between Windows and Ubuntu as a workaround, but it's very inconvenient.

[Rik]

@Sangjae They will probably not make any public statements prior to updating this answer. Any definitive answer will probably be very short term (e.g. in the release notes of a prerelease). I would not expect it for R2021b.

[Royi Avital]

@Walter Roberson, What do you mean beta for Windows on ARM. Windows on ARM was 64 Bit from day one (If I rememeber correctly). You mean x64 Bit Emulation which is in the testing. This is emulation of x64 on 64 Bit ARM (Equivalent to Rosetta 2).

[Walter Roberson]

@Royi Avital

You are correct to make a distinction between the operating system (which would potentially have been all compiled for native work) and the x64 emulation. Potentially the operating system might be more stable than the emulation.

However... the review sites that I looked at back in April all said that the operating system itself was unstable as well, even when running programs that had been compiled for ARM. It is true, though, that the x64 emulation was cited as being significantly worse than the x86 emulation.

Either way, the point stands that Microsoft's ARM implementation is notably behind Apple's ARM implementation in terms of being able to run MATLAB with any hope of stability.

[jiayi su]

It has been ONE YEAR.

[Walter Roberson]

jiayi su:

The official release date of the Apple M1 to consumers was September 17, 2020, which was the day after the R2020b release. Mathworks has had one release since then, R2021a.

R2021b is expected to be released within the next two weeks (my guess is that will be either today or tomorrow.)

[Walter Roberson]

@jiayi su as I predicted R2021b was released the same day.

[jiayi su]

@Walter Roberson But no Apple Silicon Support, either

[Stanislas Grare]

@Walter Roberson is it possible to know, for 2021b, which part is M1 native and which part is not yet ? By part I mean a toolbox, for example if I say Simscape Electrical, how will I know if it is M1 native or not ?

[Walter Roberson]

@Stanislas Grare

I am presently not aware of any method to tell which parts are M1 native.

For whatever it is worth, when I use the tool at https://github.com/DigiDNA/Silicon then no part of MATLAB is reported as being a Universal binary.

I do not know if that is the same thing as saying that nothing is implemented in Native.

[Bruno Luong]

??? Is there any benefit to deploy partial native M1 code ? Is it even possible ?

[Rik]

@Walter Roberson can you find an official date for the release? I edited the Wikipedia entries based on your comment, but I can only find pages dated 2021/09/23 with mentions of R2021b (so I'm now changing those dates).

2021/09/22 18:06 (GMT): es.mathworks.com still has R2021a

2021/09/23 11:04 (GMT): www.mathworks.com updated to R2021b

The Bing cache doesn't report the time in the header (the d= doesn't seem to be in any date format I can parse), but it mentions being from 2021/09/23 and still has R2021a on the www homepage.

[Steven Lord]

Release R2021b was released to the website the evening of September 22nd Eastern time. That could have been the morning of September 23rd in GMT.

So what you consider the "official date" for the release depends where you are. I would call it September 22nd as that's when it was released in the time zone where MathWorks HQ is located.

[Rik]

I just didn't expect it would be released in the evening of the HQ time zone. I'm a bit surprised that the only dates mentioned anywhere are 'September 2021'. I guess it is in line with how general the release date is phrased beforehand, but I find the insistance on avoiding exact dates a bit odd. Or is that just my neuroticism and completionism?

[Walter Roberson]

Mathworks does not commit to exact release dates in advance. They have internal deadlines for various aspects such as go/no-go on new features, but after that it is "as soon as all the parts are ready". That can include time to iron out priority problems identified during testing.

Documentation such as release notes and advertisements cannot be finalized until all of the go/no-go decisions are made.

Software testing usually involves assuming that all features will be released, so if any features are withdrawn then additional software testing is required.

Historically, it also included a wait period for the final software version to be pressed onto DVD, as the new version is available for physical order on day of release.

The last several years, releases have never been before 1730 Eastern time the day of release. After, that is, peak site usage for the day.

[Andrew Janke]

Totally unofficial answer here, but:

Don't expect a fully native M1/Apple Silicon Matlab build any time soon. Most of the number crunching stuff in Matlab is built on BLAS, a serious numeric library with multiple implementations. The BLAS implementation Matlab (currently) uses is the Intel Math Kernel Library ("MKL"). The ETA for an ARM port of the Intel MKL to be released is, uh, never, because MKL is a proprietary product whose whole purpose is to make Intel CPUs more useful (and thus sellable), and ARM is a competitor. (MKL doesn't even play nice with AMD x64 chips sometimes!)

So to get a full-native Apple Silicon build of Matlab working, MathWorks is going to have to do something like add support for OpenBLAS or Apple Accelerate as alternate BLAS implementation libraries, and then validate all of Matlab and all its toolboxes against that, and add compatibility shims for where the behavior of MKL and OpenBLAS differs, and then support that forever. (Plus any non-BLAS Matlab internals optimized with x64 assembly, and the External Interfaces stuff for however Python and Java and whatever work under Apple Silicon, and the Matlab Installer system which now has another degree of freedom to deal with, and probably more stuff that's not occurring to me.)

If the word "nontrivial" popped in to your head while reading that, your brain is working right!

I hope they do the work to do this, though. The M1 seems to be a kickass CPU, and well suited for numerics and scientific programming due to the high memory bandwidth of the new MacBook Pros.

[Andrew Janke]

Also I dunno if shipping Mac Matlab as Universal x64/M1 binaries is a great option? A full Matlab installation is already 10+ GB; Universalizing it would probably balloon that. My poor disk, especially for those of us who are running multiple versions of Matlab under multiple VMs on the same box.

[Michael Brown]

Thank you for an informed discussion of the issues. I just want to reinforce the observation that the M1 is an incredible processor - even with Rosetta implimentation, the memory/SSD bandwidth is beating my intel mac. I increased the size of an inverse problem on both machines with 32 GB of ram so that the the swap space balloons to 30 GB during the calculation. The intel mac slows to a crawl while the M1 soldiers on with only a modest scaling penalty from smaller problems. The major problem noted by others is the power draw when idle. I can't leave MATLAB open for long while on battery.

[cr]

I too have observed that Matlab strangly keeps processor busy drawing power even when it's pretty much idle. It doesn't happen immediately upon launch, but after I do some runs and leave it. Even clear all close all don't help and I have to relaunch Matlab.

[Wenqiang CHEN]

It's been more than a year, Do we have any progress? When the native apple silicon version will be avaliable?

[Walter Roberson]

@Wenqiang CHEN

All that is known publicly is that R2021b is not native M1 for any part of it.

What we know as MATLAB has a small number of interacting programs.

• MATLAB itself
• the help browser
• the MuPAD symbolic engine
• PolySpace
• (I think) Simulink

There might be a few others.

For Native M1 use, each of the above could in theory be converted independently, but all of each of them needs to be done at the same time. You cannot, for example, leave the Automated Driving Toolbox for later when you convert MATLAB, but you might be able to leave Simulink for later in the first version (I am not certain of this.)

Mathworks will soon be offering beta versions of the forthcoming R2022a release. I have no information about M1 status for R2022a beta. But I also would not be permitted to talk about it if I did know something, so if you are eligible for beta testing, you should watch for R2022a pre-release becoming available.

[Biraj Khanal]

Looks like I am not the only one who has been holding up to buy the M1 MBA for this sole reason :D

[Siddhant Sanjay Shah]

I think you should go ahead and buy it.. I have had no issues with the using matlab on the Macbook Air..

[Jan Theiss]

It's sad that after 21 months there is no more information than "A version of MATLAB which runs natively on Apple silicon is in development"... is native Mx support coming at all?

[Walter Roberson]

@Jan Theiss

That is a matter that is currently under Non Disclosure Agreement. If you contact your sales person, they might be willing to enter a non-disclosure agreement with you and tell you about the plans.

[Enrico Onofri]

Dec2020-March2022 is still too short for developing an Apple silicon native version of Matlab? I suspect that there are serious difficulties! E.Onofri

[Rik]

There are two main points contributing to the date at which this is ready: expected performance gain and user base.

From what I understand, the performance gain is real, but that will only show in longer/larger computations. Rosetta 2 seems fast enough to be a reasonable semi-permanent solution from what I gather.

According to this informal poll, about 13% of users use Apple products. It would be interesting to have a break-down of the people on M1, but I expect that to be <1% of the Matlab user base (a wild guess based on how long Apple has been selling ARM-powered computers).

Now, if you were a commercial company, would you drop everything else to work on converting your entire code base for a small(?) benifit for less than 1% of your customers? Or would you task a small team to do it? Obviously Mathworks chose the second option. It will be ready when it is ready. I expect Mathworks to choose a time table that makes the most commercial sense. If they publish it next year, that would be another year where the market share of Apple ARM can continue to grow, increasing the return on investment.

I don't know about serious difficulties. Maybe there are, maybe there aren't. Maybe the team working on this is too small to get it done in 15 months. Matlab is pretty big after all.

[Bowen Deng]

Here's my benchmark on my mac mini m1 with the base configuration.

With the beta 2022a with apple silicon support:

Here's the benchmark from 2021b:

[Spencer Kraisler]

The beta is fast. Using the m1 max mac studio. Ran 10 benchmarks:

Thank you to the Matlab team that developed this beta. That was a difficult task and you did a tremendous job. I hope Mathworks gives your team more developers because without add-on support, I cannot do much. 😬 Nonetheless these benchmarks show how great Matlab can work with apple silicon.

Also Ric, if I were a commerical company, I would think hard about what the future computer industry will look like. ARM is proving to be better than x86 each day, AMD is releasing better GPUs than Nvidia, and many mac users are switching to mx chips. I love Matlab, but if Matlab is too passive on staying up-to-date with the best hardware, users will inevitably switch to alternatives that do. Julia already has a great beta, and C++ and Python already have native support (even GPU). Not to mention Julia, C++, and Python are in some ways faster than Matlab, and free.

[Cye]

This is a valuable contribution. Can you please tell us which specific platform was used for these results.

[Walter Roberson]

"Using the m1 max mac studio"

[Matheus]

I have a Mac Studio M1 Max, and would like to leave these benchmarks that speak for themselves here:

[Rik]

@Spencer Kraisler

"Also Ric, if I were a commerical company, I would think hard about what the future computer industry will look like."

My name is right there. The c and the k are not close together on any keyboard layout I'm aware of.

"ARM is proving to be better than x86 each day, AMD is releasing better GPUs than Nvidia, and many mac users are switching to mx chips."

ARM CPUs tend to be more effecient and I expect the market share to increase further. NVidia is still releasing cards with better performance than AMD. So unless you're using a different definition for 'better' in these two sentences, you're simply incorrect. Many Mac users are indeed switching, but the market share is still small. If it were any other situation, I don't think Mathworks would have bothered, given a working translation layer.

"I love Matlab, but if Matlab is too passive on staying up-to-date with the best hardware, users will inevitably switch to alternatives that do. Julia already has a great beta, and C++ and Python already have native support (even GPU). Not to mention Julia, C++, and Python are in some ways faster than Matlab, and free."

Open source projects have the opportunity to be cutting edge everywhere. Mathworks needs it to be reliable. They don't need to set trends, only follow them. I don't think this thread would be the best place to discuss this matter though.

This probably isn't a good comparison, but you might want to take a look at what GNU Octave supports. There are efforts to have Octave working at all on Mac.

I don't always agree with the choices Mathworks makes (and to make it clear: I don't work for them, my only connection with them is that I'm a customer, I'm active on this forum, and have participated in some user interviews). In this case however, I completely understand the pace. In a few years when Apple-ARM gets a reasonable markte share, nobody will care anymore that it took a few releases.

[Walter Roberson]

@Rik

On the other hand, Mathworks started to lose customers because R2020a did not have full native M1 support, including GPU support. R2020a was released before the M1 series was available to anyone outside of Apple. R2020b was released the day before The Apple M1 series became available to the public, about 3 months after registered developers could get systems. People were complaining about the lack of native M1 support for MATLAB for their developer's systems.

[Agung Krisna]

How do I install Neuro Fuzzy Toolbox on the new R2022a beta? The Rosetta version is really slow to do ANFIS.

[Walter Roberson]

I am not sure if that is the same as Mathworks Fuzzy Logic Toolbox? Unfortunately none of the Mathworks toolboxes are ready for the beta as yet.

[Valeri Disko]

Installed beta version of MATLAB R2022a on my Apple MacBook Pro M1Max 64Gb RAM / OS Monterey 12.3.1. Allocated Java heap memory 12,450 Mb.

Starts much faster. UI and AppDesigner UI responds much faster. Basic operations which do not require toolboxes are fast.

Waiting for the full version to purchase.

Thank you, Mathwork team!

[Ronald Prentice]

I just ran a benchmark fractal script on this version and it's even slower than the Intel version with the Rosetta stuff. What the Hell?! Is it because I just have 8 GB of memory?

[Spencer Kraisler]

Probably. 8 gb is too small (macOS or windows) to do anything meaningful in Matlab.

[Omar Lopez]

I'd like to provide some insight into the new R2022a running in a Mac Studio 32gb Ram. Matlab is still buggy in the most basics of things. Sometimes the tool bar does not work or some windows don't close and you have to restart the program. Definatly annoying and restrics me from using Matlab at all most of the time. I find myself going to other programs that run much better on Mac M1.

[Michael Mecking]

I am running R2022a MATLAB on an M1-Mac 16 GB as well as a professor in college. My students always have a good lough about the performance of MATLAB on my Mac since I have to restart MATLAB every once in a while or MATLAB doesn't even start up after the crash. It is not annoying - it is definitely unbearable for all parties involved. I currently have no alternative but am more than willing to sacrifice the toolboxes and take a step back to work with python or octave...

When does MathWorks get the software to work on one of the most used platforms...?!

[Bruno Luong]

Do python and and octave support native M1 ?

[Paul]

TMW blog post on the subject for anyone interested.

[Rik]

Octave is hardly supported on Mac in the first place. It works through homebrew. Updates are sparse for that branch of the project, so it is much less of an alternative than on Linux or Windows.

[Joe Zhang]

@MathWorks Support Team

Currently I've installed MATLAB R2022a Rosetta 2 version and all the toolboxes. Can I use arm MATLAB R2022a at the same time, in other words, use the arm MATLAB with toolboxes running through Rosetta2?

Thanks.

[Giorgio Taricco]

I noticed that there is some confusion about the performance of M1 based Apple computers and I decided to run my own comparisons on the machines I can use. In order to simplify the comparisons, I extracted the LU factorization benchmark from the standard Matlab benchmark as the following code:

rng(1); a=randn(5200); tic; lu(a); toc

Then, I ran this code on different machines 50 times and I collected the best results:

• MacBook Pro (14" 2021 M1 Pro) Matlab R2022 U3 0.78 s
• MacBook Pro (16" 2021 M1 Max) Matlab R2022 U3 0.57 s
• iMac Pro (3.2 GHz 8-Core Intel Xeon W) Matlab R2022 U3 0.25 s
• Mac Studio (most powerful 2022 M1 Ultra) Matlab R2022 U3 0.37 s
• Mac Studio (most powerful 2022 M1 Ultra) Matlab R2022 beta 0.47 s

In view of these results, I have the following disappointing comments.

1. The best performance is achieved by the oldest 2017 Intel iMac Pro.
2. The M1-optimized beta performance is worse than the Rosetta 2 performance on the best Mac Studio.
3. The best performance with the expensive M1 Mac Studio is worse (.37 vs .27) than with the 2017 iMac Pro.

I think Apple should take in serious account all these points because many users might be tempted to avoid buying these new M1 based machines and continue using old powerhouses like the Xeon based iMac Pro.

[Patrick Dondl]

So, the beta seems to run rather well in my opinion (and with acceptable performance), but it will cease to work in a few days - where is a replacement?

[Walter Roberson]

@Patrick Dondl

There will be another, improved, beta; I have not been told the release date, though.

[Sheng Liu]

hi, @Walter Roberson since 2022b was released last week, is there any further information about the releasing date of new apple silicon beta?

[Walter Roberson]

I expect to hear dates later this week. I will ask whether I will be able to reveal them.

[법수 김]

@Walter Roberson Will 2022b beta support toolboxes and Simulink?

[Walter Roberson]

I will find out more soon

[Walter Roberson]

@법수 김

I was told "within a few weeks", nothing more precise.

[Luke OConnor]

I'm disappointed that release 2022b still does not support Apple Silicon natively, and I hope that MathWorks addresses this in 2023.

[Walter Roberson]

@Luke OConnor

Unfortunately the post from Mathworks announcing the earlier Open Beta seems to have been removed.

In that announcement, Mathworks made it clear that the Open Beta was pure MATLAB only, with no toolboxes, and that there would be additional open beta with more functionality. I am told the second beta will be available "within a few weeks".

It might be disappointing that R2022b does not support Apple Silicon natively, but Mathworks was clear in the spring that R2022b was not going to be fully native.

Back in 2021, I publicly predicted it would be R2023b before the fully native version including Polyspace was released. I am not longer confident that Polyspace will be ready by then.

[Sheng Liu]

@Walter Roberson hi~, a month has passed, is there any update on the release date for the new beta version?

[Walter Roberson]

Unfortunately I have not heard anything more.

[Paul]

New TMW Blog Post on this topic

Answer 2

[BH]

I installed 2020b on my M1 MBP today. Just thought I’d post here to let you know it runs even if it’s not officially supported

[Pierre Morel]

Could you post your best bench(5) results in Matlab on the MBP through rosetta ?

[Bruno Luong]

Can you post the result of

bench

command? Thanks.

[cr]

Viola! Appreciate you dropping an update to confirm it works. Can you please post results and screenshots of the following bench command?

>> bench(5)

What's your RAM 8GB or 16GB?

Thanks!

[BH]

8 GB RAM and 256 GB disk on R2020b Update 1 (9.9.0.1495850)

[Bruno Luong]

Anyone would bet how much gain we woud get when it runs without rosetta?

[cr]

Thanks mate.

Emulation seems to be working well on compuations but graphics are slow. Definitely looking forward to see Matlab running natively on Apple silicon.

[Wenqing Kang]

Hey BH, could you please let me know how you installed 2020b? I downloaded installer but the M1 chip Macbook Pro won't let me install, saying I need contact the administrator while I am the administrator. I believe we have the same Macbook model. Thanks.

[BH]

I didn't have any issues installing it normally so can't help you there sorry

[clappertown]

I tried Matlab R2019a on Macbook Pro 13 w/ M1 and for the same code it took 525s to finish, comparing to 456s running in Matlab R2019a on Macbook Pro 16" i9@2.4GHz.

[Shinya]

Wenquig,

This method worked for me.

I went to /Volumes/matlab_R2020b_maci64 in the terminal after mounting the image and did

# sudo open InstallForMacOSX.app

You'd have to type your password in after this command.

[cr]

Clappertown Thanks for sharing your findings. Did you check equivalency of the results from both computers?

[Richard]

Really?? Tell us please, how does the performance compare?

[Antony Ware]

I'm running 2019a, and bench produced the following scores: 0.1272 0.0836 0.0151 0.0836 3.9532 3.8510

This is overall about 60% as fast as an Intel i5 2.6GHz macbook pro, apparently.

[Amit Singh]

This video is helpful for knowing the performance of MATLAB on Macbook pro M1

https://www.youtube.com/watch?v=oNAwmu_tK7Y

[Ravi Kohli]

Thanks for the post. Does MATLAB actually work in that current (unsupported) state or is it quite buggy/glitchy?

[Ethan Woo]

I was having trouble getting R2020b to install on my M1 mac mini today, but I JUST got it!! I kept getting the "You need to contact administrator" message. I tried messing around with the terminal command (sudo open blah blah) but that didn't work. I just kept trying and trying. I also installed rosetta 2 using

/usr/sbin/softwareupdate --install-rosetta --agree-to-license

From there, I right-clicked InstallForMacOSX and selected "show package contents" - then, I navigated to Contents > MacOS. Then I double-clicked InstallForMacOSX. This brings up a terminal window that basically does nothing because it can't find a var file. So i opened another terminal window, and navigated to that crazy var file. Then, I opened the matlab_R2020b_maci64.dmg file on my desktop and double clicked the InstallForMacOSX and it actually started bouncing for the first time..

I have no idea what made it finally work. Maybe just luck... but hopefully this helps someone!

[Annie Leonhart]

All you need to do to make it work is go to "show package content" Matlab/xxxx/bin/maci64/matlabwindow.app that's pretty much it. Click that and it'll run. Nothing special you need to do. Create an alias on your dock.

Been using 2020b for weeks with 0 issues. Runs just as good as it does on PC. Fully functional.

[Ronald Prentice]

I have 2021a update 2 and it's stil very slow. Any idea when the version that will run natively on the M1 chip will be available?

[Walter Roberson]

I expect R2023b for the fully native version.

[Andrew Janke]

@Walter Roberson That's an awfully specific expectation! Do you know something I don't, or is this just a guess based on how big of a development effort this is likely to be?

[Walter Roberson]

We've danced this dance before. Polyspace is always at least 3 releases behind on technical enhancements.

Answer 3

[Moritz Kb]

It works fine on the new MBA although battery consumption is huge, meaning 3hrs from 100% to 0%.

Hopefully this will improve when it’s running natively... Also 2D and 3D times in ‘bench’ are very Slow probably because they don’t run on the GPU cores?

[Tony Davis]

The MacOS 11.1 update (out today) improved 2D and 3D 'bench' times by about a factor of 3 compared to 11.0 on my M1-based Mac Mini. Matlab 2020b Update 3 works extremely well on this setup.

[Moritz Kb]

Agreed! 2D and 3D times increased 10 and 15 times respectively from r2020b update 2 running on 11.0 to r2020b update 3 on 11.1!

[Chuck Cooper]

Is 2020b Update 3 available to all Big Sur Mac users with MATLAB maintenance, or only as a beta release?

[Walter Roberson]

It is available to everyone who is running R2020b. You should see a yellow bell near the upper right of the command window that you can click to upgrade. Or under the Help menu you can ask to check for updates.

[Walter Roberson]

Help -> Check for Updates .

The Add-On Manager will come up and list R2020b Release 3 and give you a chance to click Update. A MathWorks Update Installer will come up that has a Continue button; at that point, save any files you need and quit MATLAB (leaving the Installer open.) Then click Continue in the installer.

[Chuck Cooper]

Thanks to Walter Roberson! But when I tried installing 2020b Update 3 today, I got a message or error about not being authorized to do it. Update 3 was listed as the thing you'd get today when you download the MATLAB 2020b. (I guess that's new?) So I was not starting with a version of 2020b, as Roberson assumed. Anyhow, I chatted with Mathworks support. They are having problems with this M1 stuff and at their request I sent them a log dump of what happens when I tried to click on the "InstallMatlabForBlahBlah" item in a little window. BTW, the Mathworks tech says Roberson is amazing! But they have to vet stuff themselves.

Support said to check back in several days or a week. Boo... I suggested that perhaps as a workaround they could provide some way for a user to download 2020b update TWO and then update that to 3 as Roberson said. Mathworks said he would pass that suggestion up the chain.

[Walter Roberson]

Ah, sorry, I thought you already had Update 2.

[Chuck Cooper]

Yeah, I refrained from trying to install 2020b update 2 a few weeks ago because I thought it might cause problems later... Heh-heh. But thanks so much for your tip.

[Chuck Cooper]

SUCCESS: I did the stuff described earlier (December 4) by Ethan Woo in this thread and it worked. That included installing Rosetta (perhaps since I had never loaded any version of 2020b), and right-clicking on the package contents, etc., as described by Woo. I didn't have the problem he mentioned with the var file. It's also possible that the very simple solution suggested above by Annie Leonhart on December 8 would have worked. One other thing I did was to update Big Sur from what my Mac mini M1 shipped with a few weeks to the latest version (11.1) as of today, December 18.

[Amit Singh]

Benchmark for 2020b update 2.

https://www.youtube.com/watch?v=GGLcRF12fWU

2D and 3D scores have improved.

[Mattia Busana]

Does someone have problems with the interface? On my MB Pro M1 everything works fine, except the UI (for example Save and Open button) that is extremely laggish, I need to click 3-4 times to activate the button.

[Tommy Wilson]

For what it's worth, I'm running the new M1 chip on a 2021 MBP (Big Sur 11.6) and was also experiencing a lot of lag.

The JVM that shipped with the R2021a in my distribution was Oracle 1.8. I changed it to run on AdoptOpenJDK-8 precompiled binaries, and the lag seems to be much improved so far. See instructions on installing AdoptOpenJDK-8 via Brew here and changing/identifying Matlab's JVM here.

Note also the deprecation of:

>> brew cask install adoptopenjdk8

Which is now:

>> brew install --cask adoptopenjdk8

Hope that helps someone!

[Wenqiang CHEN]

It's been more than a year, Do we have any progress? When the native apple silicon version will be avaliable?

Answer 4

[Stanislas Grare]

Matlab 2021b on macbook m1 Air 256/16

Answer 5

[Saurabh Vyas]

I feel MATLAB Online is the way to go till we get the version built to run on Apple silicon. It's got almost everything students require and can be run on safari/chrome without an install. It also has Simulink support!!

MATLAB online

[Ji Eun Lim]

Is MATLAB online as good as the app?

[Walter Roberson]

Is MATLAB online as good as the app?

Is a Pick-Up Truck as good as a Jaguar ? Maybe not, but a Pick-Up truck is good enough for a lot of things.

[leu ryan]

But it's not useful if the data is very big to process

[Walter Roberson]

MATLAB Online is not intended as a heavy duty computing facility. It might perhaps be closer to Tuk Tuk than Jaguar. But in some cases, Tuk Tuk can be very useful to have around.

[Claudio Andres Iturra Ulloa]

Matlab online sucks prettty bad! Force me to use only python when I don't wanna take my laptop to the university, and only wanna do some work with my ipad.

Answer 6

[Ongun Palaoglu]

I am using mac mini m1, it was very good. right now it is very slow, clicking on function takes a delay. i dont understand why

[Robert Hulsey]

It takes a while because until they make full ARM support with Matlab your mac has to run an additional program (Rosetta) to translate all of the commands that Matlab gives so the computer can understand them.

[Shubam Tandel]

it just stops when i try to run any code, and always have to restart matlab.

Answer 7

[Jan van der Horst]

Anyone able to benchmark the new macbook M1 Pro\Max yet?

[Jan Valis]

Given it’s price, I’d actually love to see how it runs in the base (8 CPU, 16 GB RAM) configuration.

[Greg]

[Greg]

[Greg]

Model Name: MacBook Pro

Model Identifier: MacBookPro18,3

Chip: Apple M1 Pro

Total Number of Cores: 10 (8 performance and 2 efficiency)

Memory: 16 GB

[Zimin Lu]

I'm waiting for it too, because I find that it is a little bit slower than the macbook air (m1).

[Noel Bartlow]

I would love to see benchmark numbers for the new m1 Max if anyone has it. I have a 2019 i7 15" MBP and I get benchmark numbers above 40. Worried I'll be giving up speed if I get a new MBP.

[David Comer]

Here are my results on a MacBook Pro (16-inch 2021) with 32 GB memory abd 10 cores. The GPU(s) do not seem to be utilized in 2021b. I've attach a png of the results.

[Matheus]

@Noel Bartlow this is my M1 Max on the native build.

[Jorge Ignacio Cisneros Saldana]

@Matheus What characteristics does your Mac M1 Max has?

Answer 8

[milad karimshoushtari]

Matlab 2021b benchmarks; Left side MacBook Pro 16 inch 2019 6-core i7, and right side MacBook Pro with M1 pro 8 core.

The performance in very similar on both machines. I tried running a lot of scripts and simulink models but the performance difference was negligible.

In my opinion, matlab performance has not been improved much over the last couple of years and a natively and "Fully" supported version of matlab for apple silicon is very much necessery. And by "Fully" I mean gpu support and many toolboxes that could be supported for macos but they are only available for windows.

[Antony Ware]

Your 14" M1 Pro performance looks identical to what I'm getting on my 13" M1 MBP. And I agree very much with your opinion on what is required from Mathworks.

[Walter Roberson]

GPU support is not going to happen any time soon.

Mathworks does not write their own GPU compiler tool chain. Mathworks also does not write their own scientific math GPU library equivalent of MKL or LAPACK. Apple would have to provide those libraries and those tools.

NVIDIA invests in building these kinds of tools and in optimizing them for different architectures and models, and in providing mechanisms to optimize older versions for newer models. The mechanisms do sometimes have bugs, but when they work are able to take (for example) a Turing binary and automatically rebuild it for Ampere. These mechanisms make it practical for Mathworks to provide a single GPU implementation and have it work on a wide variety of user hardware.

Does Apple have a similar mechanism, or does Apple intend to freeze the capabilities and only add more/faster resources, or will Mathworks need to provide different GPU binaries for each different M1 capacity?

[Walter Roberson]

Next question: how fast is the Apple M1 GPU compared to the Nvidia GPUs? Mathworks' tests say that the original M1 gpu corresponds to several generations back compared to the available Nvidia models. Mathworks wonders whether anyone serious about Deep Learning would buy the M1, since much higher performance is available through Nvidia cards.

Is there even any reason to believe that STEM work is part of Apple's target market, such that they would optimize for training Deep Learning, or for execution of code heavy on linear algebra or fft? Apple advertises Video editing as the key highlight... which is a different market entirely.

[milad karimshoushtari]

I totaly agree with you about the implementations, libraries and apple's target beeing video editing.

However, I wanted to mention the M1 GPU is apple's first attempt to compete with Nvidia and in terms of performance it crushes my desktop RTX 2060 without even turining on the fans!. which is really impressive.

I think we can agree that with the current pace of apple, in a couple of years there would'nt be anyting even close to their performance in terms of both CPU and GPU.

Another thing to consider is the Neural Engine of M1, I am not sure about this but maybe Deep Learnng toolbox could also benefit from it!.

All I am saying is that Mathworks should invest more on this platform because it has the potential to become the fastest tool for engineers.

[Walter Roberson]

https://github.com/hollance/neural-engine

"Unfortunately Apple isn't giving third-party developers any guidance on how to optimize their models to take advantage of the ANE. It's mostly a process of trial-and-error to figure out what works and what doesn't."

That is a long way from Nvidia's company support for scientific computing.

Mathworks is not in the business of inventing new ways to specialize in third-party hardware that does not have a decent ecosystem. Mathworks is not in the business of doing cutting-edge AI or Deep Learning research. Mathworks is in the business of using existing third-party infrastructure to take AI and Deep Learning techniques that have matured a bit, and make the techniques easier to use for non-specialists.

Mathworks does not invest much in speculation of the "If you build it, they will come" variety. It is not going to get into the ANE ecosystem until it has reason to believe that the effort would pay off in eventual sales (either direct for that product and platform, or in terms of fame that drives sales for other Mathworks products or platforms)

Mathworks is a business. It does not generally do cool things just because they are cool: it makes market predictions and works towards them. For it to get into Deep Learning on the M1 series, it would have to be convinced that the money will be worth it.