Since upgrading to Big Sur, I experience "random" restarts of my iMac when I run Mathematica (it appears to happen only with Mathematica) after anywhere between 20 minutes and 2 hours. There is no apparent cause, other than seemingly the Big Sur version upgrade (Big Sur to Big Sur).

My system is:

iMac (27" 2020 10 Core)
Mathematica 12.3.1.0 
MacOS Big Sur 11.6 

The crash report refers to (a selection):

mdworker (process)

Exception Type:        EXC_BAD_ACCESS (SIGSEGV)
Exception Codes:       KERN_INVALID_ADDRESS at 0x0000000000000010
Exception Note:        EXC_CORPSE_NOTIFY

Termination Signal:    Segmentation fault: 11
Termination Reason:    Namespace SIGNAL, Code 0xb
Terminating Process:   exc handler [2226]

Any insights to what fix is required or similar experiences?

NB: cross posted in https://mathematica.stackexchange.com/questions/256411

I know Mathematica is not yet supported on Apple Silicon. But I am asking anyway : I have a minor issue not blocking at all and I am curious if anybody may have an idea of the problem.

I am playing with VideoGenerator. I am getting the message :

General::sysffmpeg: Using a limited version of FFmpeg. Install FFmpeg to get more complete codec support.

But I have installed FFmpeg with Homebrew.

It is installed in /opt/homebrew/bin/ffmpeg

I have a symbolic link from /usr/local/bin/ffmpeg

And the version displayed by FFmpeg is :

ffmpeg version 4.3.1 Copyright (c) 2000-2020 the FFmpeg developers built with Apple clang version 12.0.0 (clang-1200.0.32.28)

Any idea ? Should I create a separate thread ?

According to the FAQ,

https://support.wolfram.com/51740?src=mathematica

Mathematica 12.0 and later are supported on Apple Silicon and Intel on macOS 11 (Big Sur). This has been my experience.

The problem you are seeing with ffmpeg is not related. There is a support article about this somewhere, or, as suggested, you could contact support.

I just repeated it but this time with the power adapter connected and got a result of 3.07

Itai, thanks for update. Could you post here Mathematica benchmark results for all three Apple M1 machines? This information is crucial for all Mathematica users which are planning to buy Apple M1 HW during near future.

Certainly 12.0 and 12.1 will be using Rosetta. Again, we've not yet had the chance to run the full test suite with production hardware and software. If there are additional issues in running on Apple Silicon, we will do our best provide workarounds. I'm cautiously optimistic that running under Rosetta will not be significantly different than running on Big Sur Intel, but until we have the final hardware and software in hand to do a validation test, we can't give any assurances at all.

We've been working on a native port for a while now. As you note, Mathematica is a large and complex application, so the porting process is not as simple as just checking a box in Xcode. Because of the nature of the Rosetta technology, while we don't necessarily need to port every single piece at once, we do need to port some large pieces all together (for example, all LibraryLink paclets together with the Kernel). We don't have anything ready to show the outside world. Once we do, we'll release it to pre-release testers.

We are in the process of validating the 12.x releases on the Big Sur official release. When that process is finished, the FAQ will be updated.

I encountered several issues during the Big Sur beta period, some quite severe. Those issues seem to be fixed in the release candidates. Hence, I would expect 12.x to run with few to no issues on on Big Sur, but until we finish the formal validation I cannot guarantee that.

As for 12.1.1 "not being supported", that is not a typo. We simply cannot promise that existing versions of Mathematica will run a future versions of any operating system, so until we have done the validation, that combination is "not supported".

I was too curious to wait for 12.2 and upgraded to macOS Big Sur (11.0.1). I did not have any problem whatsoever with Mathematica 12.1.1. I use a iMac27inch 3.6 GHz Intel core I9. I also do not have the problem mentioned by George Woodrow and do not have the "Send Feedback to Apple" item inserted in the help menu. Maybe this was specific for the beta versions?

Those of you who just scroll to the bottom of a thread (like me)---please see the updated info in George Woodrow III's original post.

And, thank you George.

Thank you George, I’ll be careful!

The M1 chip announced yesterday only supports upto 16GB of RAM, and it is on the chip. If the 16" has the same limitation it will not be a compelling upgrade for me. I need a minimum of 32G, preferably 64G.

As various users and journalists have been noted since release, performance of Apple's M1 chip has been phenomenal. Programmers who were using an iMac Pro for builds are happy as a clam running on a MacBook Air. The processor is fast, and it rarely seems to hit thermal limits that were rampant on every Intel processor. Testers reported difficulty generating sufficient heat to turn on the fans in the M1 Macbook Pro.

The Verge's review of the M1 Mac Mini notes that there is also improved performance in swapping. In short, 16GB on these machines give far more performance than they have on other processors. Apple was clearly very frustrated with the glacial performance increases of Intel processors; they have been internally iterating on this integrated board for many years.

I'm sure performance of M1X chips -- or whatever they're named -- will have more RAM and be even better. Rumors about 14- and 16-inch MacBook Pros are starting to become very specific; I am particularly fond of Rene Ritchie's discussions on YouTube. OTOH, the 16GB version of these machines should be quite competent for many Mathematica users. I can't wait to see the benchmarks when they have a native version.