Anyone else a bit skeptical about Apple silicon future?

[smoking monkey]

iPhone 13 versus iPhone 12 performance comparisons missing - 9to5Mac

There was one interesting thing missing from yesterday's keynote: There was no iPhone 13 versus iPhone 12 performance comparison ...

9to5mac.com

If you want to feel skeptical, this article does a good job getting you to. It discusses how Apple has had "brain drain" recently in their chip designing team and how the A15 appears to have no significant performance improvements over the A14.

They jumped the gun. There are significant improvements in the A15. Check out Anandtech for a full breakdown of the chip. It's an impressive upgrade from Apple.

 

[pshufd]

What makes you think x86 binary translation is new for Windows 11? The 32-bit version has been in Windows on Arm for Windows 10 since Microsoft released their first Surface X at least. Microsoft also had 64-bit translation on Windows 10 betas but moved the release to Windows 11.

It was in Windows NT back in the 1990s too.

 

[ADGrant]

the biggest base for Apple developers come from iOS and iPadOS developers, not AppKit. Using Catalyst, it isn't a huge leap to create a macOS app if a developer already has an iPadOS app.

Even if you are an AppKit developer, recompiling for Apple Silicon is just checking a box for most developers. Ironically though, developers with a cross platform desktop applications such as Photoshop or MS Office may have more work to do because though they were cross operating system, Intel has owned the desktop for the last 15 years.

 

[Krevnik]

Ironically though, developers with a cross platform desktop applications such as Photoshop or MS Office may have more work to do because though they were cross operating system, Intel has owned the desktop for the last 15 years.

Truthfully though, a lot of that comes form the age of the codebase and features that refuse to die. Office was already ARM compatible (Android, iOS). But features like VBA were never ported to ARM for the mobile platforms, and represents low-level code that isn't as portable as the rest of it.

I do recommend listening to this episode with Erik from Microsoft if you do want to want to get a bit of a sense of the work involved with these sort of projects: https://macadmins.libsyn.com/episode-195-getting-ready-for-the-m1-with-erik-schwiebert

 

[quarkysg]

I do recommend listening to this episode with Erik from Microsoft if you do want to want to get a bit of a sense of the work involved with these sort of projects: https://macadmins.libsyn.com/episode-195-getting-ready-for-the-m1-with-erik-schwiebert

Tho. I believe there may be some truth to it (i.e. being difficult to re-develop the VBA), I think the main issue is likely from the angle of platform lock-ins. Microsoft do have a vested interest to ensure that their products works best with Windows, the same as with Apple.

 

[Krevnik]

Tho. I believe there may be some truth to it (i.e. being difficult to re-develop the VBA), I think the main issue is likely from the angle of platform lock-ins. Microsoft do have a vested interest to ensure that their products works best with Windows, the same as with Apple.

In this case, I can say that is not the issue. Yes, as Erik mentions, there are tensions, but it's nothing like what it was in the 00s.

 

[ChrisA]

There seems to be a big misconception that developers can target Intel's X86 and then have a product that runs on Mac or PC but now with Macs using M1 the developers have to do something special to make their stuff in on the Mac.

This is not the way it works. Developers have to do a LOT of special stuff to run on macOS vs Windows. It is the OS, not so much the hardware that is different.

Yes, the problem is attracting developers to the Mac. But Intel-mac and ARM-mac both have that same problem

 

[bobcomer]

Starting with Windows 11, MS has implemented their own version of Rosetta that allows x86 Windows apps to run much smoother on Windows on ARM.

I'm very aware of that, and it works, mostly. Like I said backwards compatibility is king and Microsoft knows it.

 

[dieselm]

Even if you are an AppKit developer, recompiling for Apple Silicon is just checking a box for most developers. Ironically though, developers with a cross platform desktop applications such as Photoshop or MS Office may have more work to do because though they were cross operating system, Intel has owned the desktop for the last 15 years.

Shrewdly, Apple got them to do a lot of the cross-platform heavy lifting over the last two years bringing the core of both to the massive installed base of iPhones and iPads. New subscription-based revenue models didn't hurt either.

 

[Krevnik]

Shrewdly, Apple got them to do a lot of the cross-platform heavy lifting over the last two years bringing the core of both to the massive installed base of iPhones and iPads. New subscription-based revenue models didn't hurt either.

Ironically, porting to iOS doesn’t buy you as much here as you might think.

Yes, it flushes out some of the issues that are there early, but those are generally smaller problems compared to writing a new UI layer against UIKit after writing against AppKit and something similar to Carbon before it. And any features you cut from the iOS version because they don’t make sense, or violate some App Store rule, you still need to support on the Mac.

The push to get developers 64-bit clean on macOS probably did more heavy lifting in this case. That forced these projects to grapple with the sort of issues that also happen to plague an ISA change (data alignments, etc). Apple’s changes to AppKit over the years to support 64-bit also happen to make it a lot easier for code to survive an ISA change so long as you aren’t writing custom inline assembly (which also probably isn’t as common as you might think these days).

 

[Pro Apple Silicon]

I am a complete noob and have no idea what I'm talking about in this area

Yes. Clearly.

 

[dapa0s]

Yes. Clearly.

Did you make this account just to answer to this thread? Weak.

And what are your credentials? Being an online fanboi doesn‘t really count bro

 

[leman]

The power is there, but AMD and even possibly Intel will catch up in time. I fear the future market fragmentation, with developers having to develop specifically for Apple Silicon ARM and just not having the time to do so.

I think what you are missing is the understanding that developing for an Intel CPU and developing for an ARM CPU brings very little practical difference for the vast majority of products. The big difference is developing for Windows, Linux or macOS. So if you have the resources and incentive to develop for macOS, you will develop for Apple Silicon, plain and simple. Sure, it is not always as simple as that, but on the other hand we see companies putting a lot of $$$ to build macOS versions of their software? Why do that if Mac has only 10% market share? Because Mac users are more likely to spend money on software, plain and simple. It's like Android: total world domination, but most of it's users get an Android phone because it's cheap and they want it to stay this way.

And then again, we are a year in the transition and all the major things already have native support for Apple Silicon and run great.

Yes. Clearly.

This was completely uncalled for.