M1 MacBook Air for a software engineer, tableau user and python developer

[hasanahmad]

would 8gb be good enough for this ?

 

[satcomer]

Depends if you plan to work long hours in same program so go with 16 memory sense it is it’s on the chip now!

 

[jdb8167]

As a developer, I would suggest getting the 16 GB. Tools like docker and VMs can take up a lot of RAM. It depends on what you work on but future proofing would probably be wise.

 

[hasanahmad]

As a developer, I would suggest getting the 16 GB. Tools like docker and VMs can take up a lot of RAM. It depends on what you work on but future proofing would probably be wise.

to play devils advocate. on intel machines, like Docker and VMs, Editing software would only run well on 16 gb and now they dont need that much on M1. Could the same arise when Docker and VM's become optimized for M1

 

[bill-p]

I don't know where you're getting that M1 doesn't need that much memory... I think people are misunderstanding the issue.

M1 doesn't run into the same performance penalty that past Intel MacBooks will run into on low memory <- I think that's true.
M1 doesn't need more memory <- I think that's factually false.

If you run VMs, they will always want as much memory as you can give them, so it's always better to have more.

 

[JKK photography]

Don't count on tools that needed large amounts of RAM to be performant to magically no longer need that with the M1/Apple Silicon. That's not the case. Different optimizations may make the performance hit a little less, but it's still a step change going from RAM to paging out.

Considering you mention Tableau and Python, I can also imagine that you work with large datasets. Those datasets are going to be large regardless of processor architecture/platform, not to mention Docker/VMs/etc.

You should evaluate your RAM needs independent of anyone here, but it seems likely your workflows will be best done on at least 16 GB.

 

[jdb8167]

to play devils advocate. on intel machines, like Docker and VMs, Editing software would only run well on 16 gb and now they dont need that much on M1. Could the same arise when Docker and VM's become optimized for M1

Even if the tools are currently using less RAM on the M1 do you want to chance that it stays that way for as long as you own the MacBook? Developer needs change--sometimes rapidly. Getting 16 GB means you probably don't have to think about RAM again for the life of the notebook. On the other hand, you can probably find work arounds even if you are getting low on memory in an 8 GB environment. Moving a DB container to another machine on the network for example. It depends on what you have to do.

I would echo others here though in that I doubt that you will need less RAM for the same work as on an Intel Mac. While there are probably efficiencies gained via the new Apple Silicon unified memory architecture, it is unlikely that they will be significant enough to change RAM requirements.

 

[ADGrant]

I would echo others here though in that I doubt that you will need less RAM for the same work as on an Intel Mac. While there are probably efficiencies gained via the new Apple Silicon unified memory architecture, it is unlikely that they will be significant enough to change RAM requirements.

Yes the unified memory architecture improves performance. It does not reduce RAM usage.

 

[jdb8167]

Yes the unified memory architecture improves performance. It does not reduce RAM usage.

Less replication but again, we are talking buffer not whole data sets.