I’d be really surprised if 16 gb is actually a blocker.
My iPad (same infra) runs rings around my macbook pro
4gb vs 32gb
I think we might need to rethink our perception of gb’s memory like we had to for ghz speed
3 Likes
There I have to disagree. iOS/iPadOS manage memory very differently than macOS. This isn’t a characteristic of the processor architecture, it’s a characteristic of the OS. iOS/iPadOS are absolutely brutal about limiting the amount of memory that apps can access and kicking apps in the background out of memory. Because of these limitations, well behaved iOS apps are coded to use far less memory than their Mac counterparts (and many of iOS’s limitations in terms of stuff like background app activity are downstream consequences of this).
Apple has not implemented anything like iOS’s memory limitations in Big Sur, and for good reason. Switching to an iOS memory management model would be a far more difficult challenge for Mac developers than the transition to Apple Silicon. The amount of RAM someone needs to get their work done is likely to be pretty much the same on an M1 powered Mac as it was on an Intel powered Mac.
2 Likes
I actually think this is where macOS is heading.
And that’s the point
Potential future benefit?. Yes.
Current advantages? None really that I can see.
Note Apple did not have any amazing software demonstrations at their Announcement. Perhaps because 16 Gb Unified Memory pretty much limits any practical benefit from the M1 processor?
As for “Let’s talk about this again a year from now” - I suspect we will indeed all be singing the praises of Apple Silicon. But I doubt we will be doing it with 16 Gb Memory.
Now this would all be moot if there were a way to upgrade from 16 Gb memory or if such an upgrade were anticipated into the future. But as the Unified Memory seems impossible to expand, 8 Gb and 16 Gb M1 device hardly seem like a wise choice.
Depends on your use case
I suspect for example you are not accessing a local database gigabytes in size on your iPad.
That would be nice but, as I recall, Apple started their move from PowerPC to Intel in mid 2006 and released 10.6 Snow Leopard (for Intel only) just 3 years later. I had purchased my PowerMac in late 2003 so it didn’t seem like a very long time to me.
As many have said, I agree that the fear of doom based on Apple launching pretty incredible speed and battery boosts for their low-end machines is quite perplexing.
I also think that there was nothing today to make the people who need large amounts of RAM, upgradability, 4 independent channels of Thunderbolt 3 etc, etc celebrate. But why worry? Apple haven’t said what they’re going to do with those machines yet. Would it have been better to delay these machines until the other machines were ready? Of course not.
I also think we need to see real world use before knowing if 16GB is the new 32GB for general (including ‘power-user’ use). No, it won’t work for some use-cases (hence the uncertainty for some people), but it might work amazingly for everyone else and save the earth’s limited resources in the process.
Plus, the devs in the launch video were all raving about how capable their machines were. Granted, Apple wouldn’t include them if they weren’t, but I say we should wait a bit and see what these new machines can do before predicting what they can’t.
6 Likes
I’m not buying the “because it’s Apple Silicon you only need 16GB” argument.
Particularly not if you try running iPad OS / iOS apps alongside Rosetta’d 
and native apps.
1 Like
And not Geekbench. Proper Benchmarks.
As für RAM…16GB would meet my miniumum requierements…
Another one: Parallels+Windows. Very interested in how that will work (or not).
Currently, not. While Microsoft has a version of Windows complied for ARM, they only license it to OEMs, not to individuals. So no way to get a copy of Windows running on an Apple Silicon Mac.
Parallels says: However, virtual machines are an exception. It is important to note that currently available versions of Parallels® Desktop for Mac cannot run virtual machines on Mac with Apple M1 chip. Good news: A new version of Parallels Desktop for Mac that can run on Mac with Apple M1 chip is already in active development.
So, even that might work…
With the caveat that you will likely take a performance hit since the emulation layer will not only have to run the virtual system, but also translate x86 apps for an ARM processor (since no consumer version of Windows 10 for ARM exists yet).
But with the beast that the M1 is, that might not prove such a dramatic issue (… since many Rosetta’d apps already run better on Apple Silicon than they do on native Intel chips, as the performance gap is so huge). Unless you run performance intensive apps, of course (games).
That’s also assuming that Parallels is releasing a full-blown emulator for a different instruction set.
They might just be getting their Parallels app itself to where it’s capable of running on the M1, and then anticipating Microsoft making the ARM version available in the near future.
I’ve found that Parallels is rather slippery with marketing claims sometimes, so it’s best to read what they’re actually making a hard promise to do and not read in too much.
2 Likes
Microsoft does appear to have a renewed interest in Windows on ARM (WoA) lately but I’m wondering how well applications optimized for Qualcomm chips would perform on M1 machines.
1 Like
I kind of feel like “a rising tide lifts all boats” is the metaphor here. The M1 silicon is fast, so while it may not be efficient, it would probably still perform pretty well.
The thing that concerns me, performance-wise, is if we get multiple virtualization layers involved.
Apple M1
MacOS
Parallels
Windows ARM
Windows x86 to ARM virtualization layer (supposedly being developed by Microsoft currently)
Application
Having to have the application’s stuff go through the Microsoft virtualization layer then go through the Parallels virtualization layer could potentially bork up a ton of things.
From experience, I would expect far, far fewer Windows ARM binaries to begin with - so much more of a need for this sort of setup.
1 Like
So many reasons why both M1 computers and Big Sur are essentially still beta hardware/software. Nothing wrong with that - unless you market them as the latest/greatest. And even discontinue the Intel versions (as in the Mac Air).
Not ready for prime time.
My (unfounded) feeling is that the M1’s performances are going to force Microsoft’s hand. If Apple produces blazing fast computers with insane battery life that destroys most consumer-grade Windows laptops, and Intel doesn’t catch up (which it doesn’t look like it will in the near future), Microsoft will have to look to ARM (or other alternate solutions) to catch up.
All in all, I would really not like to be Intel right now.
1 Like
What are you founding this on? Do you have an M1 laptop and do you run Big Sur?
More to the point: why do you have such one-sided hate against this platform and Apple since the beginning of this thread?
2 Likes
If the M1 computers are truly “blazing fast” then isn’t it curious that Apple did not demonstrate this or provide real-world Application metrics to prove it?
Seems highly unlikely to me that Apple would miss that opportunity.
Okay, correction: let’s revisit this thread in a few weeks when third-party benchmarks come out. I’m sure this will all be most enlightening.
In the meantime, you really do seem to have a one-sided animosity against that platform that seems more grounded in feeling that on the facts currently at our disposal.
3 Likes