Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Dannydematio

macrumors member
Original poster
May 17, 2016
52
13
Hi Guys,

I was just thinking that everyone investing in a 2020 IMac will want it working well into the new ARM (risc) future Apple is cooking up for us. But just like when Apple took us from PPC to intel, we pretty soon needed Rosetta to translate/compile for the crossover. What do you think is the best spec factor to keep the last intel iMacs running apps fast when using whatever Apple calls their emulator, especially after a few revisions and all kinds of wizardry is implemented on ARM silicon for apps?

I assume fastest processor and Ram, wins out, GPU not that important?


I am right in thinking there will eventually be a silicon emulator for the newest versions of all our apps when they go native for ARM/silicon?

Best.
 
I think you've misunderstood. As it was for Rosetta, Rosetta 2 will be used for Apple silicon Macs to run Intel apps, definitely not the other way around. Since Apple will want to incentivise developers and consumers to upgrade - and I imagine the performance penalty of using Intel to emulate Apple silicon apps would be brutal anyway.
 
If you buy a 2020 iMac, you won't need Rosetta2 at all.
It won't run on the 2020 anyway.
 
Rosetta is for AS (Apple Silicon) Macs running software for Intel hardware. But this does raise a question in my mind... what about applications that are made for Apple Silicon Macs. Will those work on our Intel-based Macs?
 
Rosetta is for AS (Apple Silicon) Macs running software for Intel hardware. But this does raise a question in my mind... what about applications that are made for Apple Silicon Macs. Will those work on our Intel-based Macs?


I pretty sure in the presentation they showed, they can compile packages to be compatible for both intel and arm and then the installer will chose the one for your architecture, If they only build for arm based macs that intel will not be able to run it.
 
I pretty sure in the presentation they showed, they can compile packages to be compatible for both intel and arm and then the installer will chose the one for your architecture, If they only build for arm based macs that intel will not be able to run it.

We have the precedent of the last transition, where I'm pretty sure most installers for Tiger and Leopard were universal binaries working both on PowerPC and Intel. Obviously then Snow Leopard dropped PPC compatibility and over time (though fairly quickly) developers stopped supporting PPC. I have a feeling though that Apple and developers will support Intel for longer this time.
 
  • Like
Reactions: pldelisle
We have the precedent of the last transition, where I'm pretty sure most installers for Tiger and Leopard were universal binaries working both on PowerPC and Intel. Obviously then Snow Leopard dropped PPC compatibility and over time (though fairly quickly) developers stopped supporting PPC. I have a feeling though that Apple and developers will support Intel for longer this time.
They don't have choice to do it. The number of Mac users is far, far more superior right now than it was back in PowerPC era.
 
  • Like
Reactions: r6mile
I think you've misunderstood. As it was for Rosetta, Rosetta 2 will be used for Apple silicon Macs to run Intel apps, definitely not the other way around. Since Apple will want to incentivise developers and consumers to upgrade - and I imagine the performance penalty of using Intel to emulate Apple silicon apps would be brutal anyway.

Actually I was aware that Rosetta is there to allow intel apps run on silicon macs. I didn't reliase it was going to be called Rosetta 2. I did confused the issue in how I expressed it. I was really thinking of 2020 macs running feature updated versions of today's important apps as time rolls on.

I guess the truth is developers will cut off intel mac users fairly quickly, (like last time) no further updates, everything will stay "as is". In a way it's probably not a great idea to buy a 2020 iMac, especially if you have say a 5k machine. Better to wait for the second iteration of silicon macs then upgrade.

I notice a lot of iMacs appear to be flooding the usual channels as people buy the 2020 models, begs the question if a stop gap machine is still a really good idea.
 
I guess the truth is developers will cut off intel mac users fairly quickly, (like last time) no further updates, everything will stay "as is". In a way it's probably not a great idea to buy a 2020 iMac, especially if you have say a 5k machine. Better to wait for the second iteration of silicon macs then upgrade.
If you want to wait 2-5 years for that, go right ahead. I, on the other hand, have work to do and the 2020 iMac chews through my workload. I'm coming from a 2012 Retina MBP.

I notice a lot of iMacs appear to be flooding the usual channels as people buy the 2020 models, begs the question if a stop gap machine is still a really good idea.
The first part of your sentence contradicts the last part. If "a lot" of people, as you say, are "flooding the usual channels" selling their older iMacs for the 2020 model, that means they clearly do not believe a "stop gap machine" is a bad idea. They just bought one!

Logic is your friend.
 
No contradiction.

Depends on how you process logic. If you're used to critical thinking you should look at both directions of the flow of thought.

What I meant and others will have understood (except you) is that since lots of people are selling their iMac 27 ( flooding the usual channels) it means some good machines are now available and thus begs the question if a stop gap machine is a really good idea, because it obviously is!

To simplify further, (just in case) lots of people are letting their current decent imac 27' late 15/16/17 and in some cases 19 go at really great prices, because they absolutely want the 2020. I've seen plenty of bargains.

So a stop gap machine is a really Good idea :)
 
Rosetta 2 shouldn't be a concern for anyone, really. As some folks mentioned, it's to allow ARM-Apples to run Intel-only applications. Additionally, LLVM (an intermediate compiler language) allows for developers to not care about processors. As long as they use a compatible compiler (like the ones with xcode) anything they write will compile for both ARM and Intel.

Recall that we also just had the No-More-32-Bit cutoff, so anything that currently runs on Intel probably has at least some support over the past few years. My guess is Rosetta 2 will mostly matter for games, where the developer doesn't care to recompile for universal binaries for ARM if they've moved on even though it *should* be fairly easy for them to do support (it should be just a checkbox and a recompile, exactly how they did it the first time, unless Apple is underselling the complexity). So I'd say the bellwether will be whether, say, older Blizzard games like SC2 and Diablo 3 get recompiled for ARM and/or how well Rosetta 2 supports them.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.