Apple Transitioning Macs to Custom Chips With 'Incredible' Performance Starting Later This Year

[Alan Wynn]

And, he keeps ignoring the point that the G5 was faster than anything Intel had at the time.

He wants to rant and ignores facts that do not support his rant. Hence why I do not respond to him any more.

Someone (AppleInsider?) did some kind of study that put at around 3%. That sounds a bit low, but I doubt a lot of people use BootCamp. What that study doesn't tell us, though, is how many people use the x86 aspect for virtualization and running Windows apps (ie. Parallels, VMWare, etc.). I'd bet that is a much bigger percentage.

Spoke to the Parallels guys a while back and was told their estimate was 8-10% run Windows for some set of applications. That was based on their own market research (and covered them, VirtualBox, and VMware). I can say that around Fox (pre-acquisition by Disney) and Paramount (this year) most of the executives had/have Macs and I can think of only one who used Parallels to run Windows (he was a heavy Excel user and liked the Windows version better). I will check with him to see if he still does it (he is not longer at Fox).

I know people doing significant 3D work on Macs.

Do not contradict the hobbyist. Never let facts get in the way of opinion.

The software aspect is a concern, especially for those of us running Windows software that isn't made at all for the Mac.

What software do you use that you need to run on Windows? (Serious question, trying to get a sense of what of the various non-macOS software will be an issue for people.)

I'd guess most of the developers making Mac apps will eventually port to Apple Silicon (but, it will probably be the more specialized pro apps that will take longer). The problem is more all the Windows apps a lot of us Mac users run currently under Boot Camp or virtualization (ie. Parallels).

Having lived through the 68K to PowerPC, the PowerPC Mac OS 9 to Mac OS X, the PowerPC to Intel, and NeXTSTEP 68K to Intel, SPARC, and PA-RISC transitions, I am pretty familiar with this process. I would bet that the closest were the various NeXTSTEP transitions. Almost none of the applications required much more than checking the switches to compile for the other architectures. Anyone who is already on Catalina has most likely done the work needed already. I expect that Adobe, Autodesk, Avid, Blackmagic Design, Maxxon and Microsoft will be shipping at release. I would also not be shocked to see a Windows solution of some sort, but would not guarantee that.

 

[firewood]

There are and were plenty of Mac users with absolutely necessary Intel, 32-bit, Carbon, PPC, 68k, 24-bit unclean, 6502, etc. apps. Apple has and will ignore and dump those people because they eventually will replace them with new, more profitable, high-end Mac customers to boost their sales and bottom line. Often better than the competition.

 

[Woochoo]

I understood your question - I don't think you understood my answer.

1. Benchmarks: For a quick back-of-the envelope comparision, I'd use Cinebench (I do 3d art, tile based rendering is critical). I would NEVER use benchmarks for a purchasing decision, however. For a purchase decision - I'd need to:

...

2. x86 Comparison: I would compare the ARM CPU to either a top of the line Rzyen 9 or a midrange Threadripper. Things the ARM System will need - 128Gb ECC ram support (as a minimum); lots of PCIe 4.0 lanes (minimum - PCIe 5.0 by the end of 2021);

4 way SMT is coming on either Zen 4 or Zen 5 - Top of the Ryzen stack will probably be a 16 core/64 thread system. Intel is going a different direction - Each core will have an Atom CPU in place of the 2nd thread. Either way, we will see a significant performance increase by end of 2021

...

Nvidia's CUDA pushes GPU computing to a new level (especially in 3d art) - Apple takes a pass.
...

Do you see a pattern here? - Can't buy a new mac pro if Timmy & Sir Idiot Boy take over 2,000 DAYS between releases.
...

My gosh dude you are trying so hard in every post to bash ARM (and specially Apple) yet you miserably fail every time. Not only that, you already started with the insults which takes away any left credibility you could have. I quoted that post but I could quote any of yours and it still wouldn't make any sense.

In the post before you were comparing a +100W high end desktop CPU (3900X which by the way consumes 77W at idle and 144W on high load, not 105W as you said) with a 5-10W tablet A12Z saying the first is more powerful. No **** sherlock, each one was designed for a really different thermal and energy envelope. You are judging ARM performance against desktops yet you haven't ever seen a consumer ARM desktop yet (maybe wait till then to compare both?). Funnily enough, the top1 supercomputer is now ARM but I guess supercomputers aren't about performance right?

Just to let you know, there are already ARM CPUs with all the stuff you said it needs to have, with 80 cores, they are the server based Ampere. That's just an insight to show how, unlike x86 that it's private between 2 companies, you can design an ARM with basically any feature you want it to have (if you have enough money). There's nothing preventing it to have those or even better specs, why is it that hard to understand?

My question is: why are you here getting mad about a transition that clearly doesn't affect you?

 

[Alan Wynn]

There are and were plenty of Mac users with absolutely necessary Intel, 32-bit, Carbon, PPC, 68k, 24-bit unclean, 6502, etc. apps. Apple has and will ignore and dump those people because they eventually will replace them with new, more profitable, high-end Mac customers to boost their sales and bottom line. Often better than the competition.

Again I ask the question: For how long should Apple maintain support for an older technology used by a small part of the user base, in a way that harms the majority of its users? People (and companies) that were dependent on apps the were not updated from Carbon to Cocoa after many years, were using software that was abandoned by its developers. The same is true for each of the other transitions you mentioned. This most recent one is a bit harder for one group, but we still do not know what will happen with them.

Let us look at three different sets of users and speculate:

1. Gamers who dual boot into Windows: The best thing that can happen for them is that Apple finally ships laptop and desktop systems that are substantially more competitive with Intel/AMD nVidia/AMD hardware and (with the addition of iPad/iPad Pro and (I hope) a new AppleTV with specs that are competitive with consoles creates a large enough market to get game developers interested. Instead of being forced to have an inferior experience on another platform, they would have a superior experience on the platform they prefer. There are signs that Apple has gotten serious about gaming (as they have about video content). In general, the previous strategy was a loser for them (anything that requires your users to run your competitor's software for a chunk of their work is bad). If this hardware transition works, it may very well improve things for this group. If it does not, it will just accelerate the trend for gamers moving off the platform for that application.
2. People dependent on Windows on Intel applications: Five possible outcomes here:
   1. The large number of ARM-based systems with seriously competitive hardware gets companies to actually port to Windows on ARM and these users migrate to that (or run their apps using Microsoft's Windows X86 emulator).
   2. A third party develops something like Rosetta but for Win32/Win64 APIs, essentially the "Mac Subsystem for Windows" (or WINE++. Translating Windows system calls to macOS systems calls (just like the Windows Subsystem for Linux does on Windows). Only the non-system call code would need to be transcoded, and if these new Macs are actually competitive, this may work fine.
   3. Apple's new systems with substantially improved price performance, coupled with the large iPad/iPad Pro market, get some of these software products to actually port to the platform or inspire a third party to create an actual macOS competitor. Either of these solutions would be an improvement over the current solution.
   4. Some of these users move to cloud or datacenter hosted versions of these Windows applications, running over something like Citrix.
   5. There is no solution for them and they leave the platform.
3. People who are dependent on Linux on Intel applications: Four possibilities here:
   1. There is a large enough base of users that makes it worth compiling for Linux on ARM to support them. For many of these applications, the per seat costs are high enough that it is certainly possible. Supporting a second instruction set on the same OS is much easier than doing a full port.
   2. The newly competitive systems get some of these vendors to port to the platform or encourage a new native competitor (just as happened with Affinity and PixelMator because of Adobe's actions).
   3. They move to cloud or data center hosted versions.
   4. They leave the platform.

The thing to recognize about the second two groups is that they must be small or the vendors would have already addressed them directly. Some of these companies already have competitors who are native, and this transition may get some users to switch tools. However, the reality is that these problems affect very few people (I would bet under 10% of Apple's market for the second two categories). If Apple is right and it can deliver what it claims, the other 90% of its users will gain substantially and that should help grow their market.

The advantage of being on Intel was that it was easy to run non-macOS Intel software. The disadvantage was that there was never going to be a way of beating the low priced clones nor the "build it yourself" crowd who do place no value on warranty support, macOS or the Apple ecosystem.This transition gives them a chance to build truly differentiated hardware. They actually have to deliver, but if they do it becomes much more interesting to develop for the platform. Previously, these niche products could say: "Well, they can just run our product under Parallels." Something I have seen from companies explaining their "macOS support". That it was so easy for them to argue that, meant they had no incentive to actually port. This transition may force them to rethink that strategy, especially if the hardware has a much better price performance profile.

 

[firewood]

Again I ask the question: For how long should Apple maintain support for an older technology used by a small part of the user base, in a way that harms the majority of its users?

What harm? Old Macs continue to work. Many old Macs (and repair parts) are available on eBay. They don't need Apple's support to keep running just those old apps if they ran them in the first place. In fact, updated OS support might actually break those old apps. Just run those old Macs (Apple II's, IBM PC's, new Dell Alienware's, etc.) behind a firewall if needed, and then buy an extra new Mac (or two or more) for any new applications.

You can still run your old but absolutely necessary 68k System 6 apps on that vintage Mac Plus (if you are part of that small user base). And still buy a spiffy new Apple Silicon Mac to put next to it.

That's actually what my desk might look like next year.

 

[Alan Wynn]

Again I ask the question: For how long should Apple maintain support for an older technology used by a small part of the user base, in a way that harms the majority of its users?

What harm?

To make it more clear, maintaining support would mean either not moving from 68K to PPC to macOS X to Intel to 64-bit only to ARM. That would harm the majority of users, to support the tiny number of those dependent on dead applications.

Old Macs continue to work. Many old Macs (and repair parts) are available on eBay. They don't need Apple's support to keep running just those old apps if they ran them in the first place. In fact, updated OS support might actually break those old apps. Just run those old Macs (Apple II's, IBM PC's, new Dell Alienware's, etc.) behind a firewall if needed, and then buy an extra new Mac (or two or more) for any new applications.

That is exactly my point. Apple should continue to move forward and you have identified one way for those people to continue to use their dead applications.

You can still run your old but absolutely necessary 68k System 6 apps on that vintage Mac Plus (if you are part of that small user base).

Yes, do that, but do not scream that Apple has screwed you by doing things that benefit most of its users.

And still buy a spiffy new Apple Silicon Mac to put next to it.

Also, look to see if there is now a competitive piece of software that has not been dead for twenty years.

That's actually what my desk might look like next year.

Running a FatMac? I still have several NeXT systems that I run for some older software and just because I like them. I know all sorts of retro-gamers who play ancient consoles and PC games. It is a hobby. My complaint is not about those people, but about the others who like King Knute stand there and shout at the tide: "Apple do not ever move forward."

 

[chucker23n1]

2. x86 Comparison: I would compare the ARM CPU to either a top of the line Rzyen 9 or a midrange Threadripper. Things the ARM System will need - 128Gb ECC ram support (as a minimum); lots of PCIe 4.0 lanes (minimum - PCIe 5.0 by the end of 2021);

Sooooo, you're not interested in the hardware, unless it has absurd specs.

Haswell Xeon Family released - Apple takes a pass.
Broadwell Xeon Family released - Apple takes a pass
Skylake Xeon Family released - Apple takes a pass
Kaby Lake Xeon Family released - Apple takes a pass
Coffee Lake Xeon Family released - Apple takes a pass
Cascade Lake Xeon Family released - Apple takes a pass

The current iMac Pro is a Skylake Xeon. The current Mac Pro is a Cascade Lake Xeon.

Do you see a pattern here? - Can't buy a new mac pro if Timmy & Sir Idiot Boy take over 2,000 DAYS between releases.

Do you really think this is appropriate?

It is only chicken-egg if a content creator is silly enough to stay with Apple:

Adobe CS - welcome to software as a service.
MS Office - welcome to software as a service.

OK, so you're not interested in spending money on software either.

You sure like complaining, though.

 

[cmaier]

Sooooo, you're not interested in the hardware, unless it has absurd specs.



The current iMac Pro is a Skylake Xeon. The current Mac pro is a Cascade Lake Xeon.



Do you really think this is appropriate?



OK, so you're not interested in spending money on software either.

You sure like complaining, though.

Apple should bend to the will of people who don't want to pay for anything ever, like all good companies do. ;-)

 

[Alan Wynn]

I understood your question - I don't think you understood my answer.

Let me try this again. You are answer a question I am not asking. The goal for questions 1, 2, and 3 is to establish an objective comparison for Apple Silicon to Intel/AMD, so that there is no moving of the goal posts by either side. This is what Jon Gruber calls “claim chowder”. Anyone will be able to look at this thread and objectively see if Apple met the goals discussed here.

I will try one last time.

1. Benchmarks: For a quick back-of-the envelope comparision, I'd use Cinebench (I do 3d art, tile based rendering is critical). I would NEVER use benchmarks for a purchasing decision, however.

None of this was about your purchase decision, it is about an objective comparison of products from two different companies. So to confirm, you are ok with Cinebench and Geekbench as the objective benchmarks? If you want to propose something else that is an objective measurement and will run on both platforms, I am open to it.

2. x86 Comparison: I would compare the ARM CPU to either a top of the line Rzyen 9 or a midrange Threadripper. Things the ARM System will need - 128Gb ECC ram support (as a minimum); lots of PCIe 4.0 lanes (minimum - PCIe 5.0 by the end of 2021);

To which Apple Silicon system should this be compared? The point of question two is that since Apple will not be selling its SoCs to others, one cannot do it purely on price of the chip, one needs some other objective metric to decide what two items should be compared.

4. Mac Computer Purchases: I have been on Mac Pros (and Power Macs before the PPC-Intel transition). You don't appear to be very familiar with the Mac Pro saga, so let be bring you up to speed.......

The point of question 4 was simply to determine Apple’s interest in each of us that answer these questions as customers. I own a several 2013 Mac Pros, a Sky Lake based iMac Pro and a Cascade Lake 2019 Mac Pro, so I am quite familiar with their product line. I have also built render farms with all sorts of hardware, and have supported many people who are professional 3D artists for Visual Effect, Motion Graphics and Games.

You may have great reasons why you did not like the products they released, however none of them matter for the question I asked.

Do you see a pattern here? - Can't buy a new mac pro if Timmy & Sir Idiot Boy take over 2,000 DAYS between releases.

You certainly come across as a very sophisticated user, I cannot imagine why Apple is not reaching out to you for your views.

2020 - Apple announces ARM Macs - Investment in 7,1 not looking like such a good idea anymore. No new software for you.

Computers are not investments, they are expenses. My B/F has already booked more than enough extra projects with his Mac Pro since January to have covered its costs. The time he would waste switching platforms would cost way more than the savings. These are tools for him, not a hobby. He wants to work with tools he likes, since he spends 8-12 hours a day using them. Finally, he will get at least 2 years out of this Mac Pro, before there is even a chance he will not see software upgrades. By that point, the machine will have paid for itself several times over.

 

[Gerdi]

1. People dependent on Windows on Intel applications: Five possible outcomes here:
   1. The large number of ARM-based systems with seriously competitive hardware gets companies to actually port to Windows on ARM and these users migrate to that (or run their apps using Microsoft's Windows X86 emulator).
   2. A third party develops something like Rosetta but for Win32/Win64 APIs, essentially the "Mac Subsystem for Windows" (or WINE++. Translating Windows system calls to macOS systems calls (just like the Windows Subsystem for Linux does on Windows). Only the non-system call code would need to be transcoded, and if these new Macs are actually competitive, this may work fine.

I am wondering why would anyone invest in 2) if 1) is readily available? Is it just to skip the Windows license cost?

And then 2) is only about translating system calls (if it is similar to WSL). There is no emulation involved in WSL.

 

[dgdosen]

...

...

Let me start by saying IMO, input from the both of you on this thread has been great. I've learned from you. I truly appreciate that and I'd suggest others look back on some of your thoughtful comments.

I am an Apple fan, and like many others here love their products. From what I heard at WWDC, I think ARM macs - especially laptops, are going to be great for the immediate future.

I have, however, much, much more skepticism regarding Apple's growing dominance and them abusing that growing dominance. I was not like this in say, 2010. To me, over the years, Apple has earned criticism. Things like decreasing repairability/upgradability on mac products; forced adoption of the touchbar; batterygate; sherlocking and other app store practices - where both US and European regulators are also taking a more skeptical view. These aren't my only reasons and I certainly don't want to debate them. I'm not persuadable on these issues at the moment (I hope I see change) and I'm not proselytizing these views onto others.

My question for the two of you - Do you you think Apple has made any materially bad mistakes in the past decade, or do you see none of that? You're been on these threads 100's of times over the last couple of days/weeks, and part of me also sees your replies as one-sided.

 

[Alan Wynn]

I am wondering why would anyone invest in 2) if 1) is readily available? Is it just to skip the Windows license cost?

Two possible explanations, depending on how it works:

1. It would provide the users a more native experience with less work for the porting company (there were companies that used Wine to ship a more native experience, since they had the source and only had to recompile rather than a full port).
2. Users would not have to worry about managing another operating system (with which they might not be as familiar), nor pay the fees, nor worry about it updates and security issues.

And then 2) is only about translating system calls (if it is similar to WSL). There is no emulation involved in WSL.

WSL was used for context, not as an exact match. You are correct, this would be halfway to Rosetta (do a binary compilation of the code, while using translated native macOS system calls).

 

[cmaier]

Let me start by saying IMO, input from the both of you on this thread has been great. I've learned from you. I truly appreciate that and I'd suggest others look back on some of your thoughtful comments.

I am an Apple fan, and like many others here love their products. From what I heard at WWDC, I think ARM macs - especially laptops, are going to be great for the immediate future.

I have, however, much, much more skepticism regarding Apple's growing dominance and them abusing that growing dominance. I was not like this in say, 2010. To me, over the years, Apple has earned criticism. Things like decreasing repairability/upgradability on mac products; forced adoption of the touchbar; batterygate; sherlocking and other app store practices - where both US and European regulators are also taking a more skeptical view. These aren't my only reasons and I certainly don't want to debate them. I'm not persuadable on these issues at the moment (I hope I see change) and I'm not proselytizing these views onto others.

My question for the two of you - Do you you think Apple has made any materially bad mistakes in the past decade, or do you see none of that? You're been on these threads 100's of times over the last couple of days/weeks, and part of me also sees your replies as one-sided.

The butterfly keyboard was a disaster, and it took them too long to acknowledge and fix it.

The touchbar is garbage. It either needs taptics (so that you have to actually press it, not just brush it, to trigger), otherwise it actually makes my experience using MBPs worse. I can’t just ignore it, because I type so fast and my fingers fly by constantly triggering things by accident. And you look at your screen, not the keyboard, so what’s the point of it?

I only like two shows on Apple TV+

I don’t care about decreasing repairability, as long as Apple can repair them. The engineering trade off is sound - smaller, lighter machines, more water tightness, etc.

Batterygate didn’t trouble me.

The App Store practices don’t trouble me because they are a minority company, not a monopoly. We see what happens if you “open up” - you get scam apps, etc. just like android.

Sherlocking doesn’t bother me, so long as Apple doesn’t use confidential information to decide how/what to Sherlock.

I think Home Pod is overpriced, and I don’t quite see the point of it.

iPad should support multiple users by now.

Apple cables break too fast.

Catalina was an awful “upgrade,” offering very little of use to users while breaking things like NAS connectivity (still not fixed).

ios 13 was a **** show for the first 6 or 7 months. The mail.app is still not right.

——
I’ve posted vociferously on many of the negative comments above here, as well as posting positive comments about other things. I call balls and strikes like I see ‘em. A lot of people are all negative all the time, or take the negativity too far.

 

[Gerdi]

Two possible explanations, depending on how it works:

1. It would provide the users a more native experience with less work for the porting company (there were companies that used Wine to ship a more native experience, since they had the source and only had to recompile rather than a full port)

Not sure how the question of porting is related at all. You either have an x86 or ARM Windows application, which you like to run. So how does porting come into play?

Regarding 2) it is clear now. You want to have something like WINE + QEMU user space emulation. This is anyway how Windows is doing it, as it only emulates user space.

 

[djjeff]

I mean everyone who has answered your questions you have not actually tried to have a proper debate with, you simply tried to discredit them. So why don’t you come with some advantages that CISC has over RISC instead, cite sources, etc like we all have done?

No one has answered my questions. That said let me ask one again: Where are the benchmarks showing the Apple desktop ARM processor faster than the x64 processors? I am not interested in any iPad benchmarks nor am I interested in any lame Geekbench benchmarks. I want to see real benchmarks of Apple desktop ARM processors obliterating x64.

Like you I don't have a vested interest one way or the other. However I think the expectations of Apples ARM processors are unrealistic. I think that's a reasonable position to take. Especially when I've heard all of this before when it came to PPC. There is absolutely nothing new here with the claims of the superiority of ARM. Maybe this time it will be different (in which case I have lost nothing) or maybe it won't (in which case I will have gained nothing). I don't have a horse in this race. I'm good either way. I just happen to think expectations are being blown way out of proportion. I have yet to see anything which demonstrates otherwise.

 

[Gerdi]

No one has answered my questions. That said let me ask one again: Where are the benchmarks showing the Apple desktop ARM processor faster than the x64 processors. I am not interested in any iPad benchmarks nor am I interested in any lame Geekbench benchmarks. I want to see real benchmarks of Apple desktop ARM processors obliterating x64.

You are asking the impossible as Apple desktop processors do not exists.
However it is possible to extrapolate the performance if you take two implementation of equal power as reference. You can also approach this from a different angle, by looking at all the deficiencies of x86 compared to ARM on architectural level.

 

[djjeff]

You are asking the impossible as Apple desktop processors do not exists.

Exactly! So any claims they're superior have to be taken with a grain of salt. Such claims are based on nothing but speculation and conjecture. Until Apple releases their ARM based Macintosh systems perhaps it would be prudent to temper performance expectations?

 

[cmaier]

Exactly! So any claims they're superior have to be taken with a grain of salt. Such claims are based on nothing but speculation and conjecture. Until Apple releases their ARM based Macintosh systems perhaps it would be prudent to temper performance expectations?

Yes they should be taken with a grain of salt. No they are not based on “nothing but speculation and conjecture.” They are based on an understanding of electrical engineering and CPU design, and the data points provided by chips that DO exist (a12, etc.), extrapolating based on Apple’s year-over-year gains and different thermal ratings, the number of cores that Apple says the new chips will have, the live demo of the a12 chip, comments by the Apple engineers about the new GPU in the WWDC statements, etc.

The lack of perfect evidence is not lack of *any* evidence.

 

[djjeff]

Yes they should be taken with a grain of salt. No they are not based on “nothing but speculation and conjecture.” They are based on an understanding of electrical engineering and CPU design, and the data points provided by chips that DO exist (a12, etc.), extrapolating based on Apple’s year-over-year gains and different thermal ratings, the number of cores that Apple says the new chips will have, the live demo of the a12 chip, comments by the Apple engineers about the new GPU in the WWDC statements, etc.

Yes, yes they are. Unless you can provide data on actual desktop ARM products anything you say is based on speculation and conjecture.

The lack of perfect evidence is not lack of *any* evidence.

Then by all means show me some evidence.

 

[cmaier]

Yes, yes they are. Unless you can provide data on actual desktop ARM products anything you say is based on speculation and conjecture.


Then by all means show me some evidence.

I just did. Everything I listed is evidence. You may dispute the weight of its probative value, but to say it’s not evidence is simply a lie.

We know (1) how fast this year’s phone and tablet chips are, and that they beat a bunch of midrange Intel silicon on various benchmarks. (2) that the year-over-year increase in performance by Apple has been much higher than Intel, so the slope of the curve is much sharper for Apple. (3) we know that these chips perform that feat using a tiny thermal solution with a very low total dynamic power. (4) simple physics says that if you take the exact same chip and stop throttling it, it will increase power but the speed will also increase. We also know you can increase the voltage to increase the frequency and increase power. So it is indisputable that as a matter of science and engineering the existing chips, if run in, say, the same thermal solution as a MBP, would run much faster than they run in a phone or ipad. (5) we saw an a12z, a two-year old chip design, running smoothly in demos. (6) we have an analyst with a fair record of accuracy predicting 50-100% performance improvement over existing intel chips in the same products (7) we have the fact that apple would be insane to make the switch if the products actually ran more slowly than existing chips (8) we have discussions during the sessions by apple engineers who instructed developers to treat the GPU as a “discrete” GPU, while noting that it is more power efficient than any AMD discrete GPU or intel’s integrated GPUs (9) we have apple taking the ray tracing/imagination tech. License (10) we know that they have been working on this chip for a long time, which is why tablets still run a12 (11) we have people who worked on the chip saying you’ll be surprised.

Those are all facts. You can dispute how much each fact is worth, but that’s not conjecture, and it’s not speculation. If you think it is, you need to go look at a dictionary, because that’s not what those words mean.

 

[djjeff]

I just did. Everything I listed is evidence. You may dispute the weight of its probative value, but to say it’s not evidence is simply a lie.

No, you did not. As was stated "Apple desktop processors do not exists". As such you cannot provide anything other than conjecture and speculation. You can argue it is informed conjecture and speculation but that doesn't change the fact that it is merely conjecture and speculation.

 

[Alan Wynn]

The butterfly keyboard was a disaster, and it took them too long to acknowledge and fix it.

Yup.

The touchbar is garbage. It either needs taptics (so that you have to actually press it, not just brush it, to trigger), otherwise it actually makes my experience using MBPs worse. I can’t just ignore it, because I type so fast and my fingers fly by constantly triggering things by accident. And you look at your screen, not the keyboard, so what’s the point of it?

I like the touch bar, (with the escape key), but I agree it needs better haptic feedback.

I only like two shows on Apple TV+

I love For All Mankind (although I think the fantasy of Teddy Kennedy beating Nixon in 1972 was silly). I like The Morning Show. While I am not the target demographic, I have enjoyed Ghostwriter (I like YA and older kids shows that do not treat kids as if they were idiots. I also like that their diverse cast does not feel like it was put together to ensure diversity, but feels more authentic). I am interested in watching (but have not yet started): See (I get it is likely to be silly, but I am curious about it), Defending Jacob, Truth be Told, Dickinson, Home Before Dark, and Servant. I have enjoyed the parts of Mythic Quest, but I do not love the style.

I am super excited about Foundation and looking forward to seeing Greyhound (I am a Navy guy so any Navy story that is not terrible is interesting to me).

I don’t care about decreasing repairability, as long as Apple can repair them. The engineering trade off is sound - smaller, lighter machines, more water tightness, etc.

I understand people’s issues here, and I would like to be able to add RAM, but I am ok with the trade off that should give me increased reliability (soldered parts should be better than parts thought a connector), as well as other benefits mentioned.

Batterygate didn’t trouble me.

The whole issue was stupid and created by people who want something about which to complain and lawyers out to cash in.

The App Store practices don’t trouble me because they are a minority company, not a monopoly. We see what happens if you “open up” - you get scam apps, etc. just like android.

Absolutely agree here. As someone who worked with a large mobile gaming company, I saw the result of Android’s open policies - almost no sales and much greater piracy. I also like the business model competition (with Android) and the fact that I do not need to go and create a million other accounts in order to use these apps/services.

I want family sharing for in-app purchases (I do not mind paying more for it, I just do not want to have to manage multiple subscriptions on my family plans). I want an optional model between subscriptions and ownership (I get to keep whatever feature set I have purchased and get to run it as along as it continues to work, even if I do not maintain a subscription. I would also prefer a more standard upgrade pricing model, but am not sure how to make that work without the “every crappy new feature is a new version” problem.

Sherlocking doesn’t bother me, so long as Apple doesn’t use confidential information to decide how/what to Sherlock.

This! If Apple cannot add any feature because some tiny third party has done some crappy implementation of an idea, we are screwed. (Even if they are great implementations, I am fine with it.)

I think Home Pod is overpriced, and I don’t quite see the point of it.

I love my HomePods (I have purchased them all for under $200 a piece), but wish they were better integrated and never grabbed the question if they could not answer it but another Siri could. Lots of other issues with them - all software/service - that I hope they fix.

iPad should support multiple users by now.

Or at least a guest mode. Related, I would love a hidden mode for notes/photos/etc. that let me hand someone my iPad without worry about them seeing confidential stuff, or handing it to a kid and having them see age inappropriate stuff.

Apple cables break too fast.

This.

Now for my primary issues:

I think Apple underestimates the value of their ecosystem and so dropped products that benefited them mostly that way. I would love to have Access Points/Routers that were tightly integrated, supported privacy and security and were best in class. I understand why they do not bother, but I think it is a mistake.

I think Apple should have built a Mac Mini Pro much earlier (discrete GPU, more Thunderbolt ports, more RAM and maybe a higher end CPU since I really want ECC). I think they should have a system/phone/laptop/etc. above the current top end with best in class performance that was expensive, but maybe at a lower margin to show what could be done (Extreme instead of Pro).

I think that Apple should not have dropped Shake, and they handled the Final Cut X transition in the worst possible way (all they would have had to do to still be the dominant player in that market would have been to say at NAB where they introduced it: ”Hey, we are releasing some really cool new updates to Final Cut 7 and plan to support it for as long as it is needed, so none of you working on projects need to worry. Today, we also want you to see something amazing. We call it Final Cut NG (for next generation). It is really the first rethinking of editing since we moved to NLEs. Play with it, give us your feedback and see how it progresses.“)

I think they should pay more attention to the Mac Mini. I think they should have a first party game controller for the AppleTV and that they should update it more frequently. (You may already have seen my suggestions on what they should do in gaming.)

I could go on, but the primary reason I do not generally bother with these points is that I know they add no value. Apple does not respond based on reading these messages, and if I want to convince them of some idea I would do it at WWDC, after a Keynote, or at a customer meeting.

 

[Alan Wynn]

Not sure how the question of porting is related at all. You either have an x86 or ARM Windows application, which you like to run. So how does porting come into play?

Let us make sure we are talking about the same thing. (I think we have lost the thread.)

I said this:

A third party develops something like Rosetta but for Win32/Win64 APIs, essentially the "Mac Subsystem for Windows" (or WINE++. Translating Windows system calls to macOS systems calls (just like the Windows Subsystem for Linux does on Windows). Only the non-system call code would need to be transcoded, and if these new Macs are actually competitive, this may work fine.

You answered this:

I am wondering why would anyone invest in 2) if 1) is readily available? Is it just to skip the Windows license cost?

To which I responded:

Two possible explanations, depending on how it works:

1. It would provide the users a more native experience with less work for the porting company (there were companies that used Wine to ship a more native experience, since they had the source and only had to recompile rather than a full port).
2. Users would not have to worry about managing another operating system (with which they might not be as familiar), nor pay the fees, nor worry about it updates and security issues.

This is all based on the idea of a third party creating the tool. Either it works as a porting kit (as the people who created WABI that got bought by Sun intended) or it acts as a user space tool as you clarify below. In the former case (the one I described in [1] - even if I did a poor job), the software company would use the tool themselves to do a port. It would be an inferior port to something that was truly native, but in many ways better than just running under a Windows VM.

Regarding 2) it is clear now. You want to have something like WINE + QEMU user space emulation. This is anyway how Windows is doing it, as it only emulates user space.

Does it all make sense now or do I still have you confused?

 

[317342]

I don't think there is anyone on this thread who could be more happy with the switch from Intel to ARM than I am.

 

[cmaier]

I don't think there is anyone on this thread who could be more happy with the switch from Intel to ARM than I am.

Do you work at Arm? (I’m guessing not because you spelled it ”ARM”)

I’m exceedingly stoked about the switch. Hoping i get a DTK.