Gene Steinberg's Mac Radio Newsletter
Issue #998
June 30, 2020
WANT TO MAKE A REAL DIFFERENCE?
Please donate to Gene Steinberg's Mac Radio! Just click here to send your payment to help the cause: Pay The Paracast Company using PayPal.Me
Thanks.
SO SHOULD YOU DELAY THE PURCHASE OF A NEW MAC?
One of the main reasons given for Apple’s decision to limit most discussion about future products is not to harm sales of existing hardware. This is certainly true for incremental product updates. So if you knew what was coming in the 2020 iPhone, you might postpone buying the current (2019) model. Well, at least when the next version is coming soon.
More or less.
But when it comes to an iPhone, Apple has been quite predictable in releasing annual updates, and for the most part, the new model isn’t so improved compared to the current model that you should rush to upgrade. At this point, most iPhone owners probably know that new models are forthcoming — and when — and they will decide if they need to wait for something special, or just buy what they want now.
When it comes to the Mac, we are very much in the same place as we were in 2005, knowing that the following year would deliver new models with Intel Inside. It came quite quickly, though. Within six months after the original announcement at the 2005 WWDC, the first MacBook Pros appeared. The last model to be upgraded, the successor to the Power Mac G5, the Mac Pro, was released in August of that year.
So certainly sales suffered, and Apple knew they suffered. But the Intel-based models were better products, and that surely fueled sales. Boot Camp and speedy Windows virtualization surely helped, then.
It’s not as important now, and Apple has confirmed that Boot Camp will be history when moving to Apple Silicon. While Intel-based virtualization apps won’t work with the Rosetta2 emulation scheme, it’s already known that there will be an ARM-based version of Parallels Desktop. That’s clear, since it was demonstrated at the WWDC, running Linux. One assumes there will be a version that all supports Windows.
But depending on the apps you run, it may take a while — if ever — for them to be ported to Apple Silicon. This means that they, too, will have to rely on Rosetta2 emulation. So the question is how much performance will you give up if you buy one of the first of the new Macs.
Well, it appears we may have an an answer, or at least a hint. Even though Apple specifically tells developers not to publish benchmarks when they get the Developer Transition Kit that is now shipping, some people aren’t paying attention.
So the first presumed published benchmarks are actually very positive. They are being run on a repurposed Mac mini, basically using the guts of the iPad Pro with the A12Z chip. Despite being run with an emulated version of Geekbench, they reveal scores that aren’t dissimilar from a regular entry-level Mac mini.
And remember, that chip is based on technology nearly two years old. It is not what Apple is expected to use when it releases the first Macs equipped with Apple Silicon, beginning, if promises are kept, by the end of this year. What this appears to mean is that most Mac user won’t notice much of a difference from emulation.
That’s quite unlike Apple’s first processor transition in 1994, where they went to the PowerPC. Emulation on the first models performed in the range of 680×0 processors two generations behind, and that meant a huge speed hit.
With the original Rosetta, used for emulation on Intel-based Macs for the first few years, there was a performance hit too, but it wasn’t nearly as noticeable. With Rosetta2, it may well be that Apple Silicon is so fast, combined with an efficient emulation system, that most of you won’t see a difference. Consider that WWDC demonstration featuring such apps as Maya, a powerful 3D rendering app, which performed well enough to make it seem that it was running native.
Now that doesn’t mean all Intel apps will run. Kernel extensions will have to be updated, and, beyond virtualization software, there may be apps that are heavily dependent on old code bases and thus need lots of work to upgrade. Apple claims that most developers can probably deliver Apple Silicon updates to their apps with a few clicks and a recompile of Xcode, and that may be true.
But there are going to be apps that may never be updated, or are highly complex to modernize, and thus will never make the transition, or at least not for a few years.
It’s promising that Apple has worked with Adobe and Microsoft to make sure that many of their apps will be run native on Apple Silicon from Day One, as will Apple’s Final Cut Pro X and Logic Pro. Indeed, Apple’s video editing software can evidently do amazing things on Apple Silicon, running multiple 4K streams and performing updates almost instantaneously. At least that’s what was demonstrated at the WWDC.
So for most of you, it may very well be worth waiting to buy that new Mac. Depending on the apps you run, you may find that there will be little or no performance or compatibility hit.
But what if you really need a new Mac now, and you’re not inclined to wait for up to two years for the model you want to transition to Apple Silicon?
Let me give you a quick and dirty observation.
I’ve just run some quick Geekbench 5 benchmarks on my Late 2014 27-inch iMac. I ran it rough, without a fresh restart, to get a fast comparison with current hardware. Comparing it to the current model, labeled Early 2019, surprised me considerably. Single core benchmarks are no better than 25% faster. Multi-core scores are twice as fast because there are twice as many processing cores.
So for most of the things I do, I highly doubt that the current iMac would be worth the expense. So long as my computer continues to run well, I can continue to use it for a while, and wait for the Apple Silicon version.
Now Apple promises that it will continue to release Intel-based Macs, and that it will continue to provide macOS upgrades. Obviously they aren’t saying for how long, but I expect moving to Apple Silicon will be done as quickly as possible, even before the mid-2022 deadline. OS upgrades will probably be available for at least fix years, based on current practice.
So macOS Big Sur will run on Macs up to six or seven years old. I see little reason to expect much of a change with its successors.
That returns us to the core question, which is whether sales of new Macs are destined to collapse as Apple works on the new hardware. I wouldn’t even begin to predict, and, as far as Apple is concerned, it doesn’t matter. Apple has fully baked in the consequences and benefits of its WWDC announcements. Unlike many tech companies, it has long-range plans, and does its best to fulfill them.
But those benchmarks explain fully why Apple made the move. When I consider the really modest performance improvements of the iMac over five years, I can see the logic in Apple’s decision to go with its own chips. Remember that Apple claims a 100-fold performance boast in Apple’s A-series chips since 2010. Imagine realizing anything near that as Macs with Apple Silicon receive ongoing upgrades.
In the meantime, if my iMac continues to behave itself, I probably have a decent amount of time to save up for my next computer. I hope nothing happens to change that decision.
THE FINAL WORD
The Tech Night Owl Newsletter is a weekly information service of Making The Impossible.
Publisher/Editor: Gene Steinberg
Managing Editor: Grayson Steinberg
Marketing and Public Relations: Barbara Kaplan
Copyright 1999-2020 Making The Impossible. All Rights Reserved.
Privacy Policy: Your personal information is safe with us. We will positively never give out your name and/or e-mail address to anybody else, and that's a promise!
Issue #998
June 30, 2020
WANT TO MAKE A REAL DIFFERENCE?
Please donate to Gene Steinberg's Mac Radio! Just click here to send your payment to help the cause: Pay The Paracast Company using PayPal.Me
Thanks.
SO SHOULD YOU DELAY THE PURCHASE OF A NEW MAC?
One of the main reasons given for Apple’s decision to limit most discussion about future products is not to harm sales of existing hardware. This is certainly true for incremental product updates. So if you knew what was coming in the 2020 iPhone, you might postpone buying the current (2019) model. Well, at least when the next version is coming soon.
More or less.
But when it comes to an iPhone, Apple has been quite predictable in releasing annual updates, and for the most part, the new model isn’t so improved compared to the current model that you should rush to upgrade. At this point, most iPhone owners probably know that new models are forthcoming — and when — and they will decide if they need to wait for something special, or just buy what they want now.
When it comes to the Mac, we are very much in the same place as we were in 2005, knowing that the following year would deliver new models with Intel Inside. It came quite quickly, though. Within six months after the original announcement at the 2005 WWDC, the first MacBook Pros appeared. The last model to be upgraded, the successor to the Power Mac G5, the Mac Pro, was released in August of that year.
So certainly sales suffered, and Apple knew they suffered. But the Intel-based models were better products, and that surely fueled sales. Boot Camp and speedy Windows virtualization surely helped, then.
It’s not as important now, and Apple has confirmed that Boot Camp will be history when moving to Apple Silicon. While Intel-based virtualization apps won’t work with the Rosetta2 emulation scheme, it’s already known that there will be an ARM-based version of Parallels Desktop. That’s clear, since it was demonstrated at the WWDC, running Linux. One assumes there will be a version that all supports Windows.
But depending on the apps you run, it may take a while — if ever — for them to be ported to Apple Silicon. This means that they, too, will have to rely on Rosetta2 emulation. So the question is how much performance will you give up if you buy one of the first of the new Macs.
Well, it appears we may have an an answer, or at least a hint. Even though Apple specifically tells developers not to publish benchmarks when they get the Developer Transition Kit that is now shipping, some people aren’t paying attention.
So the first presumed published benchmarks are actually very positive. They are being run on a repurposed Mac mini, basically using the guts of the iPad Pro with the A12Z chip. Despite being run with an emulated version of Geekbench, they reveal scores that aren’t dissimilar from a regular entry-level Mac mini.
And remember, that chip is based on technology nearly two years old. It is not what Apple is expected to use when it releases the first Macs equipped with Apple Silicon, beginning, if promises are kept, by the end of this year. What this appears to mean is that most Mac user won’t notice much of a difference from emulation.
That’s quite unlike Apple’s first processor transition in 1994, where they went to the PowerPC. Emulation on the first models performed in the range of 680×0 processors two generations behind, and that meant a huge speed hit.
With the original Rosetta, used for emulation on Intel-based Macs for the first few years, there was a performance hit too, but it wasn’t nearly as noticeable. With Rosetta2, it may well be that Apple Silicon is so fast, combined with an efficient emulation system, that most of you won’t see a difference. Consider that WWDC demonstration featuring such apps as Maya, a powerful 3D rendering app, which performed well enough to make it seem that it was running native.
Now that doesn’t mean all Intel apps will run. Kernel extensions will have to be updated, and, beyond virtualization software, there may be apps that are heavily dependent on old code bases and thus need lots of work to upgrade. Apple claims that most developers can probably deliver Apple Silicon updates to their apps with a few clicks and a recompile of Xcode, and that may be true.
But there are going to be apps that may never be updated, or are highly complex to modernize, and thus will never make the transition, or at least not for a few years.
It’s promising that Apple has worked with Adobe and Microsoft to make sure that many of their apps will be run native on Apple Silicon from Day One, as will Apple’s Final Cut Pro X and Logic Pro. Indeed, Apple’s video editing software can evidently do amazing things on Apple Silicon, running multiple 4K streams and performing updates almost instantaneously. At least that’s what was demonstrated at the WWDC.
So for most of you, it may very well be worth waiting to buy that new Mac. Depending on the apps you run, you may find that there will be little or no performance or compatibility hit.
But what if you really need a new Mac now, and you’re not inclined to wait for up to two years for the model you want to transition to Apple Silicon?
Let me give you a quick and dirty observation.
I’ve just run some quick Geekbench 5 benchmarks on my Late 2014 27-inch iMac. I ran it rough, without a fresh restart, to get a fast comparison with current hardware. Comparing it to the current model, labeled Early 2019, surprised me considerably. Single core benchmarks are no better than 25% faster. Multi-core scores are twice as fast because there are twice as many processing cores.
So for most of the things I do, I highly doubt that the current iMac would be worth the expense. So long as my computer continues to run well, I can continue to use it for a while, and wait for the Apple Silicon version.
Now Apple promises that it will continue to release Intel-based Macs, and that it will continue to provide macOS upgrades. Obviously they aren’t saying for how long, but I expect moving to Apple Silicon will be done as quickly as possible, even before the mid-2022 deadline. OS upgrades will probably be available for at least fix years, based on current practice.
So macOS Big Sur will run on Macs up to six or seven years old. I see little reason to expect much of a change with its successors.
That returns us to the core question, which is whether sales of new Macs are destined to collapse as Apple works on the new hardware. I wouldn’t even begin to predict, and, as far as Apple is concerned, it doesn’t matter. Apple has fully baked in the consequences and benefits of its WWDC announcements. Unlike many tech companies, it has long-range plans, and does its best to fulfill them.
But those benchmarks explain fully why Apple made the move. When I consider the really modest performance improvements of the iMac over five years, I can see the logic in Apple’s decision to go with its own chips. Remember that Apple claims a 100-fold performance boast in Apple’s A-series chips since 2010. Imagine realizing anything near that as Macs with Apple Silicon receive ongoing upgrades.
In the meantime, if my iMac continues to behave itself, I probably have a decent amount of time to save up for my next computer. I hope nothing happens to change that decision.
THE FINAL WORD
The Tech Night Owl Newsletter is a weekly information service of Making The Impossible.
Publisher/Editor: Gene Steinberg
Managing Editor: Grayson Steinberg
Marketing and Public Relations: Barbara Kaplan
Copyright 1999-2020 Making The Impossible. All Rights Reserved.
Privacy Policy: Your personal information is safe with us. We will positively never give out your name and/or e-mail address to anybody else, and that's a promise!