Leaving Apple over ARM Based Macs discussion [merged]

[gnasher729]

For those who do professional work, would anyone consider switching to Microsoft Surface Book for productivity if Apple decides to go full ARM with their professional lineup?

That would be absolute nonsense to do.
[automerge]1592923967[/automerge]

Apple has clearly stated that Rosetta will NOT translate VM, so full windows would require some third party to do a huge work.

That's news to me. Where did you find that? Tim Cook did _demonstrate_ it on stage.
[automerge]1592924078[/automerge]

My question is if the bit-level format of a REAL, DOUBLE, etc as stored in a binary file will be the same. Will they follow the same IEEE formatting standards that are used on Intel hardware. I ran into this problem when moving from SunOS SPARC to Linux Intel. All binary data files had to be reformatted at the word level. I'm not talking about image files, but data files.

Intel and ARM use same byte ordering, same alignment rules. No need to touch any files.
[automerge]1592924195[/automerge]

It won't simply because windows runs on intel, and Mac will not in the future.

And there is Rosetta 2. For software that runs on Intel.
[automerge]1592924292[/automerge]

Hardcore gamers, no. But the mobile gaming market is far bigger than the PC gaming market.

They are just not that loud.

 

[jerryk]

Is this video available online, or is it only for WWDC attendees?

The WWDC keynote is on youtube. The information about the transition to Apple Silicon is in the last 20 minutes or so.

 

[rhor]

That's news to me. Where did you find that? Tim Cook did _demonstrate_ it on stage.
[automerge]1592924078[/automerge]

What Tim demonstrated was that ARM Macs can virtualize ARM OS's. Rosetta won't translate VMs, which is clearly stated here.

 

[gnasher729]

I left the apple ecosystem for 3 reasons;

-The newer iPhones are incredibly hard to repair compared to the older ones.
-A move to ARM will obsolete my older equipment faster and I will not be able to buy new at this time, or anytime in the foreseeable future.
-Unsure about Hackintosh support.

Taking the last line: Do you want me to give an estimate how much Apple cares about that?
[automerge]1592924611[/automerge]

What Tim demonstrated was that ARM Macs can virtualize ARM OS's. Rosetta won't translate VMs, which is clearly stated here.

We'll see. Rosetta doesn't translate virtualisation instructions. That doesn't mean Apple can't create an ARM VM that runs plain non-virtualised code, like Windows. Running VMs on Windows would then fail.
[automerge]1592924721[/automerge]

I am stunned that Apple never considered AMD chips in the equation. It's not like Intel is the only game in town ya know?

I'm sure they considered three alternatives: Stay with Intel, switch to AMD, switch to ARM. And then they made a decision.
[automerge]1592924831[/automerge]

ARM will go to the MacBook Air or 12” MacBook.

ARM will go to all MacBooks, iMacs and Mac minis, with the exception of the iMac Pro with 12 or more cores initially.

 

[gnasher729]

ARM does not blow away X86 out of the water. It is simply a reduced instruction set which allows it to have way less transistors than X86. That is how it is able to achieve such a low power for mobile devices. X86 has complex instructions which are not needed for simple tasks that mobile devices do, so that is why ARM is more suitable in an iPhone or iPad (as the most complex X86 instruction set will not be used and cause extra battery drain).

For devices were complex tasks are done, that is were ARM will start to struggle as it requires much more operations than a X86 processor. X86 also has much more compatibility and is able to optimize incoming instructions.

Oh please. You are completely confusing what a processor does with what a user sees. ARM has the same vector instructions that Intel has. Intel has some complex instructions for encryption/decryption which an iPhone handles more efficiently with specialised hardware. Intel has complex string instructions which nobody uses. The Apple implementation of ARM beats Intel in instruction decode, available processing units (up to nine instructions executing simultaneously), simultaneous dispatch, and instruction retirement. And Apple's ARM processor comes with a powerful AI unit, where Intel has nothing.
[automerge]1592925608[/automerge]

ARM solves the Hackintosh problem. They can make custom ARM chips with their own instructions or variants so that their OS will only run on their ARM chips. I want to see what they come up with. If they can provide x86 Translation (not emulation), that runs 99% of software, that will be good enough for me.

But "Hackintosh" is not a problem for Apple. Individuals building their own machine running MacOS are so few that Apple ignores them. There was _one_ company that tried to sell PCs running MacOS as a business, and they found themselves ordered to pay $2,500 per PC shipped by the courts. (Not that Apple got any money from that, nor did the company's lawyers, because you can't take money out of an empty pocket). But Hackintoshes, Apple can just ignore. And of course they will keep running for many years, until all new PCs come with newer Intel processors that are too new to be supported by the current MacOS. And by then PCs with ARM chips might be widely available and we have a new generation of ARM Hackintoshes.

 

[naturalstar]

Ok, I wanted to clarify understanding before I fully responded because it has been stated over and over that the entire Mac line is transitioning to ARM. Nothing will be left on Intel. I’m curious as to what will transition first. Apple has 3 new ARM chips ready to go. Where is the 12-core processor going? That’s not going into an MBA or MB. 12 Cores to browse the net and post on social media? That smells of either an iMac/iMac Pro or maybe even the MBP 16” instead. To add, Prosser also stated on a podcast last month that the 16” MBP was going ARM early according to his sources. Right or wrong, I believe Apple is going to let everyone know they are serious about the transition by being swift and by throwing in a “professional” machine transitioning earlier in the overall process. Starting with 2021, each new release will be ARM-based (unless COVID continues to affect timelines).

As far as readiness, I’d agree, but Apple historically makes swift, user experience affecting changes unapologetically and leaves consumers with a take it or leave it decision to make. Being that I have a 10th gen 13” MBP now, my concern lies in how long will it be supported and what the ultimate benefits will be for ME in the ARM transition (I understand how Apple is benefiting). The order of which products transition to ARM and how fast are less concerning to me because either way, by the time I’m in the market again for my next Mac, everything will be ARM.

To answer OP’s question, overall I have spent more time with Windows than Mac so returning wouldn’t be a problem. I do expect that my current MBP will be supported at least for the next 3 years minimum. By that time, if Apple ends Intel support, then I’ll evaluate where my workflow stands and what’s available to meet it. I wouldn’t get a Surface Book since there’s better Windows-based laptops out there, but I get what you were asking in general.

I'll just quote myself from earlier because a lot of what I stated pre-WWDC remains the same post-WWDC. I should be good for the next 3 years which is what I wanted. Still wondering about the ability to run Windows-based software in Parallels post-transition, but it ultimately comes down to really wanting a single machine that I could use on-the-go as a matter of convenience. I'll keep watching as this transition unfolds.


Just to add to some of the conversation, I watched the PSTOU address and there were some telling points that underpinned the Keynote.

1) There was a friendly admonishment that the first ARM Mac will be ready later this year, so developers will only have a few months to get their app(s) ready.
2) Apple will have the CPU, GPU, and add a neural engine for ML apps. That's a pretty big point that was glossed over in the Keynote. AMD doesn't have long as well, it would seem.
3) Apple has been working with Unity and there was a demo
4) There was a mention of being able to boot from an external drive because they are committed to working with other OS platforms and no disruption to existing development work - still no specific mention of Windows, but Linux was mentioned.

Take that as you will.

 

[CmdrLaForge]

IF someone needs on a daily basis to use Windows as an OS I think it might make sense to leave the Apple ECO system. IF not than I think for someone like me as an Apple products only user everything on Mac will just get a lot better. I am really excited about this switch!

 

[leman]

I am stunned that Apple never considered AMD chips in the equation. It's not like Intel is the only game in town ya know?

Not much difference between AMD and Intel at this point. That Apple decided to take a huge risk and play their own game is an indication that they are VERY confident in their ability to deliver high performance.

And it might just pay off - I mean, their iPad GPU is not that far off a mid-range dedicated Nvidia GPU. I wouldn’t be surprised if their first ARM 13” will match the 5300M from AMD.

 

[matram]

2) Apple will have the CPU, GPU, and add a neural engine for ML apps. That's a pretty big point that was glossed over in the Keynote. AMD doesn't have long as well, it would seem.

I think the question about the GPU is an interesting one.

Does the fact that they will start with a 13" MBP mean that they are confident in that they have more than enough CPU capacity to beat Intel, and that they can beat the Intel integrated GPU but not a dedicated GPU?

The question is then what happens in the next stage. Will Apple be apple to do their own GPU that would beat an AMD dGPU or will they continue to use AMD for the 16".

 

[kazmac]

As some in this forum know (rather well), I’ve been trying to leave Apple since 2017.

Have had a ridiculous amount of trouble with Apple hardware and software these past 3 years. I hung in as I wasn’t ready to make the leap, but job requirements and this announcement cements my decision to finally leave.

 

[matram]

As some in this forum know (rather well), I’ve been trying to leave Apple since 2017.

There is nothing saying that it has to be a binary decision. Even if I could no longer use a Windows VM I would still buy a MBP but I maybe also buy a Thinkpad for the very few things where I really need Windows.

I run a fairly complex windows app I developed myself for professional use. Knowing that this was in the pipe I started porting that from WinForms / C# to SwiftUI some time ago. A very interesting and rewarding exercise. I do really like what Apple is doing!

 

[kazmac]

There is nothing saying that it has to be a binary decision. Even if I could no longer use a Windows VM I would still buy a MBP but I maybe also buy a Thinkpad for the very few things where I really need Windows.

I run a fairly complex windows app I developed myself for professional use. Knowing that this was in the pipe I started porting that from WinForms / C# to SwiftUI some time ago. A very interesting and rewarding exercise. I do really like what Apple is doing!

I tried to stay dual platform, but it just is not working for my job, and what I want to do for fun.

 

[SocialKonstruct]

I would find it hilarious if Apple replaced Intel and started to license their own CPUs for Windows 10 based machines

 

[leman]

Does the fact that they will start with a 13" MBP mean that they are confident in that they have more than enough CPU capacity to beat Intel, and that they can beat the Intel integrated GPU but not a dedicated GPU?

To expand on my last post (just above yours), the 3DMark Ice Storm benchmark score is 221182 for the A12Z (current iPad Pro GPU). The Ice Storm for the 5500M Pro is 364184, about 1.6 times faster. The Ice Storm for The Iris Plus G7 (current Ice Lake 13" 10-gen) is 146308. And wile comparing these results is not the most accurate thing (old benchmark, different OS etc. etc.), it does give us some indication to what these GPUs can do — and I think it is not bad at all given that we are talking about the iPad.

Now, take the same SoC, give it a 30W TDP (rather than what I'd guess is 10W in the iPad), and you have something that can compete with lower-mid-range dedicated gaming GPUs out there (like Nvidia 1650). Or double the amount of cores, and you get a 2070 RTX-level performance at under 40W TDP — assuming it scales as well as Apple seems to imply.

 

[matram]

To expand on my last post (just above yours), the 3DMark Ice Storm benchmark score is 221182 for the A12Z (current iPad Pro GPU). The Ice Storm for the 5500M Pro is 364184, about 1.6 times faster. The Ice Storm for The Iris Plus G7 (current Ice Lake 13" 10-gen) is 146308. And wile comparing these results is not the most accurate thing (old benchmark, different OS etc. etc.), it does give us some indication to what these GPUs can do — and I think it is not bad at all given that we are talking about the iPad.

Now, take the same SoC, give it a 30W TDP (rather than what I'd guess is 10W in the iPad), and you have something that can compete with lower-mid-range dedicated gaming GPUs out there (like Nvidia 1650). Or double the amount of cores, and you get a 2070 RTX-level performance at under 40W TDP — assuming it scales as well as Apple seems to imply.

Very interesting, thank you for elaborating.

 

[Useless Touchbar]

To expand on my last post (just above yours), the 3DMark Ice Storm benchmark score is 221182 for the A12Z (current iPad Pro GPU). The Ice Storm for the 5500M Pro is 364184, about 1.6 times faster. The Ice Storm for The Iris Plus G7 (current Ice Lake 13" 10-gen) is 146308. And wile comparing these results is not the most accurate thing (old benchmark, different OS etc. etc.), it does give us some indication to what these GPUs can do — and I think it is not bad at all given that we are talking about the iPad.

Now, take the same SoC, give it a 30W TDP (rather than what I'd guess is 10W in the iPad), and you have something that can compete with lower-mid-range dedicated gaming GPUs out there (like Nvidia 1650). Or double the amount of cores, and you get a 2070 RTX-level performance at under 40W TDP — assuming it scales as well as Apple seems to imply.

If it really works that way and you could have gaming laptop level performance (or even better) on a cool, efficient, slim and sleek MacBook in the future, why wouldn't game developers create their games for ARM Mac as well in the future?

 

[AndyMacAndMic]

If it really works that way and you could have gaming laptop level performance (or even better) on a cool, efficient, slim and sleek MacBook in the future, why wouldn't game developers create their games for ARM Mac as well in the future?

Computer speed is only a small part of the equation. Game developers/companies create their games for those places in the market where they sell the most, because AAA games are very expensive to make.

For AAA games the biggest part of the market is consoles (PlayStation, XBox) and the PC. Mac computers occupy less than 10% of the market world wide. How much procent of that 10% are gamers who want to play AAA games on a Mac? Probably a fraction of that.

It is very costly to convert those games to another platform. The new Mac can be the best and/or fastest gaming machine in the world, but in the end it is a simple cost-profit analysis.

 

[Useless Touchbar]

Computer speed has nothing to do with it. Game developers/companies create their games for those places in the market where they sell the most because AAA games are very expensive to make.

For AAA games the biggest part of the market is consoles (PlayStation, XBox) and the PC. Mac computers occupy less than 10% of the market world wide. How much procent of that 10% are gamers who want to play AAA games on a Mac? Probably a fraction of that.

It is very costly to convert those games to another platform. The new Mac can be the best and/or fastest gaming machine in the world, but in the end it is a simple cost-profit analysis.

Of course. But if slim and power efficient MacBooks suddenly become the best performing laptops on the planet then surely that would turn some heads on the PC side and make them think about switching? Game developers go where gamers go, and as far as I know, gamers go where the performance is. Apple might never compete on performance per dollar with Windows machines but superior performance overall should still mean something for gamers and developers alike.

 

[bill-p]

Of course. But if slim and power efficient MacBooks suddenly become the best performing laptops on the planet then surely that would turn some heads on the PC side and make them think about switching? Game developers go where gamers go, and as far as I know, gamers go where the performance is. Apple might never compete on performance per dollar with Windows machines but superior performance overall should still mean something for gamers and developers alike.

Nope. The Nintendo Switch is definitely not the most powerful gaming system around.

Neither was the Wii.

Marketshare is everything. The Mac doesn't and won't ever sell well enough to be taken seriously. Apple just put the Mac into a coffin in terms of gaming.

 

[AndyMacAndMic]

Of course. But if slim and power efficient MacBooks suddenly become the best performing laptops on the planet then surely that would turn some heads on the PC side and make them think about switching?

Of course, but those are a lot of ifs. And as I said: the speed of a computer is only a very minor part of the equation. It is not easy to turn around a well established market. And 'turning some heads' as you say, how many heads would that be to turn around an entire market?

Also Apple themselves never were focused on creating and optimizing Macs for AAA games. It is simply not their shtick. Apple knows it has lost that game (excuse the pun) a long time ago. Why would they focus on that in the future? The portable game market (for iOS tablets and iPhones) is huge. Those games can be sold via the Apple (games) store and also run on a Mac ARM. Apple can make a profit here and will probably focus on that. Can they make a profit on AAA games on a Mac? Not likely (if the Mac stays an open platform).

 

[AxiomaticRubric]

Is this video available online, or is it only for WWDC attendees?

Anyone can watch the WWDC videos after logging in with a free Apple Developer account.

 

[luvbug]

For those who do professional work, would anyone consider switching to Microsoft Surface Book for productivity if Apple decides to go full ARM with their professional lineup?

Apple made it crystal clear yesterday that they will be selling Intel-based Macs "for years to come", so what's the question, again?

 

[sub150]

Nope. The Nintendo Switch is definitely not the most powerful gaming system around.

Neither was the Wii.

Marketshare is everything. The Mac doesn't and won't ever sell well enough to be taken seriously. Apple just put the Mac into a coffin in terms of gaming.

I predict Apple will buy a game studio or two and start making their own real AAA games for Mac/AppleTV/iPad.

It'd be interesting if the next AppleTV had TB3/USB4 and you could hook two together to increase your performance for gaming, with MacOS 11.1 combining the two A chips on the fly.

 

[LeeW]

If you need a new MBP or whatever, buy one. Sure change is coming, but for those that are actually going to use one, get one, you will get your money's worth out of it.

If you are one of these, buy now, sell now, wait for a year type, then you clearly don't really need one so do whatever, just do it quietly and without having to post about it

 

[StellarVixen]

I think that wise choice would be first to wait and see.

I am a bit sad that Apple is leaving x86, but it might actually turn out to be good.

Let’s just wait.