Some of the Fortran libraries are used in parts of the Pangeo ecosystem, and I figured it would be useful to share this thread on installing gfortran on Apple M1 chips from the fortran discourse here in case anyone is running into issues!
Side note - it is awesome to see a similar forum (https://fortran-lang.discourse.group) to this!
Thanks for sharing. Is anyone familiar with a “master” thread anywhere that tracks potential issues with running the scientific Python stack on M1/M2 computers? I’ve definitely put off on a much-needed laptop upgrade while I wait for stuff to shake out, but it’s hard to keep track of where folks are continuing to run into issues with the core compiled libraries that accelerate the stack.
I haven’t paid too much close attention, but I have an M1 Mac and I use miniforge’s osx-arm64 distribution and haven’t run into any issues. Most of my heavy lifting happens on a Linux compute cluster, but I’ve never noticed any problem running scripts/notebooks with the scientific python stack on my local machine when I have.
Should also mention that Conda also recently added M1 support/builds: Anaconda | New Release: Anaconda Distribution Now Supporting M1
@darothen that is a great question! I have heard users expressing issues with using xESMF (really the ESMF part) on M1 chips, but have not tried it myself yet as I still own an intel chip Mac.
Awesome, thanks @mgrover1 and @lsterzinger. It seems like there’s pretty solid ground. Apple also has a 30-day refund policy in case things go awry 