MacInTouch Amazon link...

competition, technology, Apple management & plans

Channels
Apple, Products
I had a "gee whiz, that's fast!" experience the other day when i booted my 2010 Mac Pro 5,1 into Snow Leopard. Fast, fast, fast! And that's on a spinning drive.
Many years ago, I timed the boot-to-usable-desktop time on my Macintosh II: 5 seconds. That was on a spinner, too! Although for full disclosure, I should state I had all extensions disabled. I think the normal boot time was around 15-20 seconds. I was doing packaging graphic design with Freehand at the time.
 



Ric Ford

MacInTouch
It looks like Apple may see more competition for its lucrative "wearables" business and the data that comes with it:
Reuters said:
Google takes on wearables giants with $2.1 billion Fitbit deal
Alphabet Inc-owned Google will buy fitness tracker pioneer Fitbit Inc for $2.1 billion, as the search giant looks to take on Apple and Samsung in the fast-growing market for wearable devices.

... Google, which has been defending its privacy practices after a number of regulatory probes, said it would be transparent about the data it collects for its wearables.
#privacy #security
 


Ric Ford

MacInTouch
More management changes at Apple:
Mark Gurman said:
Apple Names Five VPs, Including Return of Early iPhone Executive
Apple Inc. recently promoted several executives to vice president, a key title at the company reserved for the most influential players.

In the past month, the company named Paul Meade a vice president of hardware engineering, Jon Andrews a vice president of software engineering, Gary Geaves to a new vice president of acoustics role, and Kaiann Drance as a vice president of marketing.

The Cupertino, California-based technology giant also brought back Bob Borchers, a former iPhone executive who recently worked at Google and Dolby Laboratories Inc. Borchers is now a vice president of marketing.
 


Ric Ford

MacInTouch
Even as Apple apparently prioritizes marketing and sales over development of innovative, high-quality products, some of its latest changes hit its longterm advertising partner hard:
Mark Gurman said:
Apple Ad Agency Cuts 50 Jobs as iPhone Maker’s Needs Evolve
Apple Inc.’s outside advertising agency has cut about 50 employees, according to people familiar with the matter. The firm said it was adjusting to the changing needs of its only client.
 


[FYI:]
Ars Technica said:
Google Pixel 4 review—Overpriced, uncompetitive, and out of touch
It's the fourth generation now, yet we've got to ask—what's the point of the Pixel line?

The Pixel 4 arrived on the market as one of the most leaked, most talked about smartphones of 2019. This year, Google seems like it is really trying to find something unique to offer, with new features like the Google-developed "Motion Sense" radar gesture system, face unlock, a 90Hz display, the next-gen Google Assistant, and a new astrophotography mode.

At the prices Google is asking, though, the Pixel 4 is hard to recommend. The company saddled the phone with an ultra-premium price tag, but the Pixel 4 can't compete with ultra-premium phones. The phone falls down on a lot of the basics, like battery life, storage speed, design, and more. The new additions like face unlock and Motion Sense just don't work well. It seems like Google just cut too many corners this year.
 


Ric Ford

MacInTouch
It's pretty obvious that Apple has become a media company as it now develops software for platforms that compete with its own computers:
Neowin said:
Apple may be working on a successor to iTunes on Windows 10
... A job listing on LinkedIn looking for a senior software engineer says, "Join us and build the next generation of media apps for Windows."
Later in the description, it also says, "If you love music and you are passionate about writing code, and want to work with world-class engineering teams that ship to millions of users, the Media Apps team is the place for you." Naturally, this implies that an Apple Music app could be on the way, but it seems likely that if that's the case, then an Apple TV app shouldn't be too far behind.
And, like Google and Facebook, Apple is looking for billions of dollars in the advertising business:
Reuters said:
Apple could raise annual ad income to $11 billion by 2025: JPMorgan
The launch of Apple TV+, coupled with Apple Inc’s foray into digital services, could help the company increase its income from advertising by more than five fold to $11 billion annually within the next six years, analysts from JP Morgan said on Friday.
 


Ric Ford

MacInTouch
Intel confirmed some new competition for Apple's Mac Mini though macOS doesn't support these computers and prices haven't been announced yet:
AnandTech said:
Intel Confirms Comet Lake-Based NUC 10 ‘Frost Canyon’ UCFF PCs
... The Intel NUC 10 platform will have a Thunderbolt 3 port controlled by Intel’s Titan Ridge chip, USB 3.2 Gen 2 and USB 2.0 Type-A ports, GbE, HDMI, and the usual audio connectors. As an added bonus, the Frost Canyon NUC PCs are also equipped with far-field microphones supporting Amazon’s Alexa and Microsoft’s Cortana assistants.

With rather powerful processors and sophisticated connectivity, Intel’s Frost Canyon UCFF PCs look very potent for various applications except gaming as Intel’s UHD Graphics can barely satisfy those who play demanding titles. Good news is that the systems feature a Thunderbolt 3 port that can be used to connect an external graphics box, but the latter tend to be rather expensive.

Intel’s NUC 10 will be available in the coming weeks. Prices have not yet been published.
 


Intel confirmed some new competition for Apple's Mac Mini though macOS doesn't support these computers and prices haven't been announced yet:
I had to smile when I saw this. I've been thinking about this item:


Apple's excellent software QC (specifically the macOS 10.15.1 update) seems to have killed my eGPU setup. The NUC is less than half the price I paid for my late 2018 Mini. Same SSD and RAM [but] only one Thunderbolt 3 – enough for a direct connection to the Razer Core X.

Apple 'discussion' follows my increasing level of frustration. (All told, I reinstalled Catalina three times.)

If I didn't despise Microsoft since its behavior in the '90's; and/or felt more joy in the past running the hobby OS that is Linux (almost as much trouble as the current macOS), I'd jump tomorrow.

#applequality
 


Ric Ford

MacInTouch
Here are Apple's latest financial results...
And for comparison...
CNBC said:
HP rises on earnings beat
... HP's biggest business segment, Personal Systems, which includes sales of notebook PCs and desktops, contributed $10.43 billion in revenue, up 4% year over year. That beat the $10.27 billion consensus among analysts surveyed by FactSet.
Reuters said:
Dell misses revenue estimates on weakness in server business
... Client Solutions Group, which includes desktop PCs, notebooks and tablets, reported a 4.6% rise in revenue to $11.41 billion.
 


if there's one thing Tim Cook has shown over the years, it's that what's best for the customer isn't even on Apple's radar any more.
I appreciate your kvetching, and I'm right there with you emotionally. However....

Cook's fiduciary duty is to the shareholders of the corporation. That is immutable, and really shouldn't be a source of angst.

The question is whether paying no attention to "what's best for the customer" begins to bite the corporation's bottom line on behalf of the shareholders. Right now Apple investors are being rewarded by the current strategies. Clearly, the kind of computers and performance that we appreciated from Apple are trending away from our preferences toward other markets. Are those markets happy with their subscriptions? Love their watches? Earbuds comfortable? Are they pleased with their keyboards? Or not? Will the customers rise up and punish the corporation for failing them? Will revenues fall? That will get the shareholders attention, and it will result in changes. Probably not back to what Apple computer hardware was once known for, though.

But right now we're not seeing any rebellion in the marketplace. I don't think "we here" are considered to be Apple's best customers anymore, so Apple is not trying to do what's best for the customer we represent. As long as the shareholders remain happy, I wouldn't expect that to change.
 




Ric Ford

MacInTouch


Ric Ford

MacInTouch
Don't let Apple tell you it isn't in the advertising business and that you're not a product being profiled, packaged and sold for profit as part of its "services" business....
And, like Google and Facebook, Apple is looking for billions of dollars in the advertising business....
A bit more on the topic:
The Street said:
How Far Can Apple Stock Go? Far, Says J.P. Morgan
... "While investors are trying to identify the next big frontier for services, we believe hidden in plain sight and underappreciated by most is the advertising opportunity within Apple's fingertips, given the secular migration of advertising dollars to mobile platforms, the large installed base of close to (1 billion) iPhone users, and importantly, Apple's successful exploration of advertising to date," Chatterjee wrote in his Nov. 15 research note.
#advertising
 


This sounds a lot more like a 2013 Mac Pro gone to plaid than a return to the actual cheese grater paradigm. There has to be room for a product between the iMac Pro equivalent of a BMW and the 2019 Mac Pro equivalent of a cargo jet.
Apple should have updated the the 2013 Mac Pro to become the non-pro 2020 Mac. Drop Xeon processors and go with single Intel HEDT socketed processor, only one PCIe GPU, two M.2 slots on the motherboard and Thunderbolt 3/USB-C, and a lower price.

A single GPU and CPU wouldn't overwhelm the current cooling solution.

Allow choices of silver chrome finish and project RED. Start it at $2999.
 


Ric Ford

MacInTouch
Apple should have updated the the 2013 Mac Pro to become the non-pro 2020 Mac. Drop Xeon processors and go with single Intel HEDT socketed processor, only one PCIe GPU, two M.2 slots on the motherboard and Thunderbolt 3/USB-C, and a lower price.
A single GPU and CPU wouldn't overwhelm the current cooling solution.
Allow choices of silver chrome finish and project RED. Start it at $2999.
I think virtually everyone here would agree that we'd like to buy a product like that at a price like that (or a little lower), while that sort of price would still leave Apple plenty of profit margin (given the components) and would sell many millions of computers.

Here's the problem: If Apple did that, it would kill sales of its other pricy products. So, it won't.

(Actually, considering the alternatives, $2999 is probably way too high.)
 


I think virtually everyone here would agree that we'd like to buy a product like that at a price like that (or a little lower), while that sort of price would still leave Apple plenty of profit margin (given the components) and would sell many millions of computers. Here's the problem: If Apple did that, it would kill sales of its other pricy products. So, it won't.
(Actually, considering the alternatives, $2999 is probably way too high.)
I considered the pricing, as well, and although the market alternatives would be cheaper, a fully configured Mac Mini runs around $3200, so I'm guessing a market does exist at this price.

You are correct in thinking that it would take away sales of the Mac Mini, iMac Pro and Mac Pro, but shouldn't really matter to Apple which product a person buys, as long as it's Apple? High margins for this product make it more attractive for Apple to offer this, and it's easier to later drop pricing (to $2499) than to raise it.
 


Apple should have updated the the 2013 Mac Pro to become the non-pro 2020 Mac. Drop Xeon processors and go with single Intel HEDT socketed processor, only one PCIe GPU, two M.2 slots on the motherboard and Thunderbolt 3/USB-C, and a lower price. A single GPU and CPU wouldn't overwhelm the current cooling solution. Allow choices of silver chrome finish and project RED. Start it at $2999.
I think you just described the current Mac Mini – small, single-processor, limited GPU, limited storage, Thunderbolt 3 for expansion.
 


I think you just described the current Mac Mini – small, single-processor, limited GPU, limited storage, Thunderbolt 3 for expansion.
Indeed, the 2013 Mac Pro is basically a Mac Mini on steroids. The only major difference besides Xeon and ECC RAM is that the Mac Pro has some GPU options, and more (if inadequate) cooling. If Apple were to improve the current Mac Mini's cooling system and added an option for a discrete GPU, that would basically be an iMac without the screen. Make Xeon an option and it's an iMac Pro without the screen. Maybe that's getting too close for their desired market segmentation. I think they'd need to jump a bit higher to make a new product work, but then you're getting more into X-Mac territory, which would be nice, but they're clearly not interested in that.

What was discussed in last night's Accidental Tech Podcast recording was the cheese grater (Mac Pro and Power Mac G5) and even the previous G4 and G3 tower designs, where they put lesser components in the same big case. Yes the 2019 Mac Pro has an expensive case, but it's hundreds of dollars, not thousands. Same with the double-sided motherboard. Those Xeon processors are quite pricey, even if they're not as much as the dual-socket variants. So Apple could make a Mac Pro that looks the same as the current one, but with a simpler single-sided motherboard for i7 or i9 processors, fewer PCI slots (none of which are MPX modules), a smaller power supply, maybe even one less fan, and all they'd have to do is blank out some of the PCI covers, if that. Previous Mac Pro and G5 iterations had single socket and dual socket setups in the same case after all. This could get down closer to the $4,000 or $3,000 range, especially with ARM CPUs.
 




Ric Ford

MacInTouch
Previous Mac Pro and G5 iterations had single socket and dual socket setups in the same case after all. This could get down closer to the $4,000 or $3,000 range, especially with ARM CPUs.
Did you mean to say AMD CPUs?
Nope, but that would have a similar (though lesser) effect.
Wow, really? I mean, we know that macOS runs on AMD systems (as Quinn Nelson demonstrated), but, as far as I know, neither macOS nor any actual Mac apps run on ARM hardware, so your imagined system wouldn't be a Mac but something entirely different (and irrelevant to the Mac Pro discussion). That's essentially the same thing as running macOS and Mac apps on an upgraded Raspberry Pi.

(Yes, I know that Apple ported a few of its apps from its ARM-based iOS platform to macOS via the buggy Catalyst emulator, but that's also something entirely different.)
 


It's pretty much a forgone conclusion at this point that Apple is going to switch macOS to ARM at some point in the not-too-distant future. It's just a question of exactly when and how that transition is going to take place. Intel has been dropping the ball nearly as badly as IBM and Motorola were in the PowerPC era, and I don't think Apple wants to take the risk on AMD and potentially end up in a similar situation in a couple more years. iPhones already beat the 2019 Mac Pro in single-core performance, and Apple can get ARM chips much more cheaply than if they had to buy them from Intel or AMD.

I have no doubt there's already a version of macOS for ARM, just like there was an Intel version of Mac OS X for its first five versions before the Marklar project was finally revealed at the 2005 WWDC keynote. One of the points Steve Jobs made in his address was that the Mac OS had to be processor-independent, making such transitions easier. Maybe this isn't quite the same, but since these are also Apple's chips, they have a much better understanding of how to port the OS, or even to modify the chips themselves to do what they want.
 


A Raspberry Pi isn't even close to an Apple ARM package. Going ARM wouldn't be a huge issue at all performance wise, especially if there weren't the power constraints that phones and tablets have: 64 ARM cores in a desktop would be pretty easy, and going higher wouldn't be tough at all. And given how long Apple had an Intel build of OS X, it would be naive to think that there isn't an ARM build floating around. The performance wouldn't be any handicap, and Catalyst's issues are as a bridge technology – going full ARM would simplify things greatly.

GeekBench 4 results:

Raspberry Pi 4​
Single: 978​
Multi: 1768​
iPad Pro (A12X)​
Single: 5026​
Multi: 18040​
Xeon E-2176M (mobile) 2.7GHz, 6 core​
Single: 5232​
Multi: 19199​
MacBook Pro (13in, 2019, i7)​
Single: 5469​
Multi: 19092​
 


Ric Ford

MacInTouch
It's pretty much a forgone conclusion at this point that Apple is going to switch macOS to ARM at some point in the not-too-distant future...
And how do you think that's going to work, exactly? We'll boot up the ARM-Mac from our Carbon Copy Cloner backup and start running Mac apps on it?

Or, we'll buy an ARM-Mac and it'll connect over WiFi and import our documents and apps from an existing Mac and start running them?

Or something else?

Will developers have to create new apps for the ARM-Mac that we'll have to buy anew?

Will there be an emulator, like Rosetta? Will our apps be slow?

Do you think there will be any compatibility problems?

Certainly, we can run macOS and Mac apps in a VM now, and, if Apple didn't restrict it, a VM could run on pretty much any hardware, I guess. Do you think the ARM-Mac will run macOS and Mac apps in a VM?

How soon do you think we'll see this "foregone conclusion" ARM-Mac?
 


And how do you think that's going to work, exactly? We'll boot up the ARM-Mac from our Carbon Copy Cloner backup and start running Mac apps on it?

Or, we'll buy an ARM-Mac and it'll connect over WiFi and import our documents and apps from an existing Mac and start running them?

Or something else?

Will developers have to create new apps for the ARM-Mac that we'll have to buy anew?

Will there be an emulator, like Rosetta? Will our apps be slow?

Do you think there will be any compatibility problems?

Certainly, we can run macOS and Mac apps in a VM now, and, if Apple didn't restrict it, a VM could run on pretty much any hardware, I guess. Do you think the ARM-Mac will run macOS and Mac apps in a VM?

How soon do you think we'll see this "foregone conclusion" ARM-Mac?
Probably it won’t be much more difficult than it was when we transitioned to PowerPC or to Intel. Apple has had two runs at this whole process, and unlike those times, they have actually been compiling software for the intended platform for years.

And so have all of the software publishers. Office and Creative Cloud will both be mostly there by the end of next year. And those subscriptions already cover both iOS and macOS (and Windows and Android, too).

Microsoft has been working on the same thing (x64 on ARM), as well; emulation isn’t a tough nut to crack anymore (we’ve got emulators for nearly every CPU platform ever made at this point).

And Ubuntu and other Linux distros already run on ARM. You can buy an ARM workstation now. So, while it probably hasn’t been done publicly (the count of ARM workstations that would need to run macOS right now is probably less than my fingers), but Apple could have a build of QEMU or something similar going (again, just like they did with x86 ports).

And throwing a ton of threads at a problem isn’t the worst thing. It allows the machine to work around slow emulation by not bogging down everything. (That was one of of the amazing parts of the PowerPC transition: there wasn’t a ton of CPU to work with, and the OS wasn’t a highly compartmentalized and highly threaded piece of code.)

Some MIPS CPUs have some extra instructions specifically to boost x64 emulation. If Apple controls the whole stack, getting some crucial x64 compatibility instructions included wouldn’t be difficult at all.

Connecting over WiFi to copy over data is Migration Assistant. Why wouldn’t that be an option?

If I recall correctly, we were able to have boot drives that could start both Quadra (68K) and Power Mac (PowerPC) models.
 


Ric Ford

MacInTouch
Here's some relevant background for ARM-Mac discussion:
Microsoft said:
...
What limitations should I be aware of when running a Windows 10 ARM-based PC?
There are some limitations when you run a Windows 10 ARM-based PC:
  • Drivers for hardware, games and apps will only work if they're designed for a Windows 10 ARM-based PC. For more info, check with the hardware manufacturer or the organization that developed the driver. Drivers are software programs that communicate with hardware devices—they're commonly used for antivirus and antimalware software, printing or PDF software, assistive technologies, CD and DVD utilities, and virtualization software.
    If a driver doesn’t work, the app or hardware that relies on it won’t work either (at least not fully). Peripherals and devices only work if the drivers they depend on are built into Windows 10, or if the hardware developer has released ARM64 drivers for the device.
  • 64-bit (x64) apps won’t work. You'll need 64-bit (ARM64) apps, 32-bit (ARM32) apps, or 32-bit (x86) apps. You can usually find 32-bit (x86) versions of apps, but some app developers only offer 64-bit (x64) apps.
  • Certain games won’t work. Games and apps won't work if they use a version of OpenGL greater than 1.1, or if they rely on "anti-cheat" drivers that haven't been made for Windows 10 ARM-based PCs. Check with your game publisher to see if a game will work.
  • Apps that customize the Windows experience might have problems. This includes some input method editors (IMEs), assistive technologies, and cloud storage apps. The organization that develops the app determines whether their app will work on a Windows 10 ARM-based PC.
  • Some third-party antivirus software can’t be installed. You won't be able to install some third-party antivirus software on a Windows 10 ARM-based PC. However, Windows Security will help keep you safe for the supported lifetime of your Windows 10 device.
  • Windows Fax and Scan isn’t available. This feature isn’t available on a Windows 10 ARM-based PC.
And here's some more:
Engadget said:
Surface Pro X review: Gorgeous hardware marred by buggy software
The Surface Pro X is the thinnest Surface yet, with the slimmest bezels I've ever seen on a Microsoft tablet. Cosmetically, this thing is a major upgrade, but the Surface Pro X is also different on the inside. It uses a custom chipset that Microsoft says is the fastest ARM processor in a PC and has an integrated gigabit LTE radio for speedy cellular data. The Surface Pro X is Microsoft's latest attempt at making Windows run on an ARM-powered device, after the mess that was the Surface RT.

... The Surface Pro X is a beautiful piece of hardware and the best Snapdragon-powered PC around. It offers a great 2-in-1 experience, a speedy gigabit LTE connection and an excellent keyboard case (for an extra $140). But the entire Windows on Snapdragon platform is plagued by limited app compatibility and bugs. If you must have Microsoft’s software to get work done on the go, the Surface Pro X is a solid choice. But most people will be better served by either a Chromebook or the Surface Pro 7.
This might be relevant, too:
The Verge said:
Adobe deals with ‘painful’ early reviews of Photoshop for iPad
At the kickoff keynote for Adobe Max, the company’s massive annual creativity conference, 15,000 designers and creatives cheered as Photoshop on the iPad was unveiled onstage. The long-anticipated app had been teased since last year’s conference, and the air in the Los Angeles Convention Center was filled with excitement as attendees finally got to try it out in between workshop sessions and panels hosted by inspirational speakers. But online was a different story, as negative reviews poured in on Twitter and YouTube, confirming early reports that the app was missing key features and felt unfinished.
(As far as I know, there aren't any hackintoshes running on ARM. Please let us know if anyone can find some - or even macOS and Mac apps running in a VM on ARM.)
 



It's pretty much a forgone conclusion at this point that Apple is going to switch macOS to ARM at some point in the not-too-distant future. ...
I'm not sure it's a foregone conclusion, but I agree. I predicted this switch would happen years ago. It is a logical conclusion of bringing chip design in house and not relying on large foundries. It may not seem cost-effective up-front, since macOS and apps will need re-writing, but in the long term it gives Apple far greater control over their products and destiny.

With each successive macOS, Apple has been moving the Mac to a closed environment like the iPhone/iPad. The vast majority of devices plugged into Thunderbolt, USB, or DisplayPort don't care what the OS is. Special devices will need special drivers, that's always been the case.

I think the new Mac Pro will remain Intel for the foreseeable future. As we've discussed, its target market is not most users. Apple will have to have two versions of macOS to support it though. They did this for Mac OS X Tiger, so it's not a feat they couldn't repeat.
... And given how long Apple had an Intel build of OS X, it would be naive to think that there isn't an ARM build floating around. ...
I agree. Obviously that means Apple would already have some ARM hardware running macOS. Modified iPad Pros could be a starting point.
Will there be an emulator, like Rosetta?
I would hope so, but given the new Apple mindset, I wouldn't assume so. If there's no emulator, Apple may work with larger software vendors, like Microsoft and Adobe, to make sure ARM versions are available when the ARM Macs debut. They may let smaller developers flounder for tools and documentation like they have with APFS and Catalina.
 


Ric Ford

MacInTouch
The vast majority of devices plugged into Thunderbolt, USB, or DisplayPort don't care what the OS is.
That's only because the OS includes software to support them, all of which Apple would have to rewrite for ARM. But we can see that happening now as Apple gradually adds more external device support in iOS updates.
 


That's only because the OS includes software to support them, all of which Apple would have to rewrite for ARM. But we can see that happening now as Apple gradually adds more external device support in iOS updates.
Right, just like the Intel transition, the OS, drivers, and Apple software will be ready upon rollout. No doubt they're already maintaining parallel code bases. That could even explain some of the infuriating iOS-ification of apps over the years, like Photos, Mail, Final Cut(?), and now all the apps iTunes was split into. Their changes may not just be due to maintaining feature parity with their iOS counterparts, but also the underlying code that has to be ARM64 compatible.

For the rollout question, I think there are two ways they could go about it. First is to be aggressive and quickly transition their entire product line, if not the Mac Pro. To do that though, they'd need an emulation/compatibility layer. Whether that manifests like the 68K -> PowerPC transition or PowerPC -> Intel, I have no way of knowing, but the beauty of both of those strategies is that they were completely seamless to the end user. Under the hood, my guess is that it would work more like Rosetta, but for the end user it would be more like 68K -> PowerPC, where the PowerPC processors were so much faster than their predecessors that the older code still ran markedly better than on native hardware.

One of the clever things about Rosetta is that on multi-core systems it would split the code translation and code execution into different threads/cores, thus improving performance. That sort of paradigm could potentially be dialed up significantly with 16+ core ARM Macs to ensure that the execution instructions aren't waiting for the translation instructions to finish, or vice versa. At the time of the Intel transition though, a lot of Macs were just dual-core, so Rosetta translation wasn't as fast as it could've been.

The other way Apple could go about it, without doing an emulation layer, is to more gradually roll out ARM to the lower-end products, like the MacBook Air and Mac Mini, and then work their way up to the iMacs, MacBook Pros, and eventually the Mac Pro. There could also be parallel releases of the same product just with the different architectures, but I think this whole idea is already getting very confusing for the consumer.
 


Speaking of ARM-powered systems, about two weeks ago I decided to check out the Surface Pro X, so I purchased a "bundle" from Microsoft (that included a Surface type cover and surface pen). It arrived a couple of days later.

After unpacking it and turning the device on, I noted with glee the gorgeous screen – very iPad Pro-like. The entire system looks really nice.

So off I went to begin the installation of the applications I need. First up, Cisco's AnyConnect. Oops, wouldn't install. Next up, Sophos Anti-Virus. Oops, wouldn't install.

I went through the iteration of "oops, wouldn't install" about a dozen or more times.
And the software packages I was able to install were installed as 32-bit apps in emulation mode. Thus, those apps ran dog slow. Even Microsoft Office ran like molasses in the Arctic.

There is absolutely no 64-bit support for apps running on the Surface Pro X, until such time as developers port their software over to ARM.

So, I packaged the Surface Pro X back up and requested a refund from Microsoft. The Surface Pro X is a nice-looking device, for sure. But until such time as the ARM-based processor in the device is fully supported by Microsoft and app developers, it's merely a nice-looking paper weight.

I'm sure glad I held onto my Surface Pro 7. Although it's outdated in design, it is a speedy little 2-in-1.
 


It's pretty much a forgone conclusion at this point that Apple is going to switch macOS to ARM at some point in the not-too-distant future. It's just a question of exactly when and how that transition is going to take place. Intel has been dropping the ball nearly as badly as IBM and Motorola were in the PowerPC era, and I don't think Apple wants to take the risk on AMD and potentially end up in a similar situation in a couple more years. iPhones already beat the 2019 Mac Pro in single-core performance, and Apple can get ARM chips much more cheaply than if they had to buy them from Intel or AMD.
I have no doubt there's already a version of macOS for ARM, just like there was an Intel version of Mac OS X for its first five versions before the Marklar project was finally revealed at the 2005 WWDC keynote. One of the points Steve Jobs made in his address was that the Mac OS had to be processor-independent, making such transitions easier. Maybe this isn't quite the same, but since these are also Apple's chips, they have a much better understanding of how to port the OS, or even to modify the chips themselves to do what they want.
Are we, then, to wait for all the macOS developers out there to port their software from x64 to ARM? Might be a long, long wait.
 


That's only because the OS includes software to support them, all of which Apple would have to rewrite for ARM. But we can see that happening now as Apple gradually adds more external device support in iOS updates.
Yes, Ric, Apple would have to rewrite for ARM. But would every developer who writes software for macOS? This would be a lengthy process for some, and some developers would probably not even bother.
 


I'll look forward to seeing the ARM-Mac editing and transcoding raw 8K footage on a dual Pro Display XDR setup then.
At the high end, you make a very relevant point, especially if having a macOS device doing it.
But for consumer devices, it isn't that big of a stretch. Apple seems to show middling interest in the high-end workstation market, and as a product line, it makes me wonder if there is a future there. Porting is hard work, but how hard depends on the quality of the code you have.

It isn't all that hard to see iOS devices gain even more Mac-like capabilities and, in different configurations, become the future 'desktop'.

But if there isn't a real future for pro desktops, yet removal of the line is of concern because of indirect impact on sales of other product lines, then doing an update every few years with very high cost makes sense. You don't have consumer-level marketing and stocking requirements. So the token effort that we seem to see looks like that sort of very long, very slow strangulation of a product line that is 'maturing out' of relevance.
 



Given Apple's philosophy of dragging everyone forward whether they like it or not, I think an ARM option is inevitable, too.

What is to be gained by switching from x86 to ARM? For the end user, possibly more efficient power savings and battery life, so laptops that run longer on a single charge. Also, in theory, possibly cheaper computers, but Apple has been loathe to drop pricing generally.

For Apple, there's a lot of work involved, but the gains may be worth it. They wouldn't have to pay Intel for CPUs, and possibly not even AMD for GPUs, since all of the IP [intellectual property] would be available in-house. Those cost savings would mean higher margins for Apple.

It doesn't have to be an all or none transition, in the beginning at least. Lower-end units and laptops might be where things are ARM-only, and Mac Pros and up might still be x86.

Of course, Apple also then retains control of CPU releases and the architecture. And there are potential downsides too, including incompatibility, stability, legacy software etc.

ARM designs are making inroads outside of embedded and handhelds, since modern CPUs are mostly heat-limited in performance:
ElectronicDesign said:
Server Chip Startup Ampere Gains Big Name Backer
... Intel holds more than 95% of the global market share for server chips and its x86 architecture —which is also used by AMD—is the gold standard in data centers. Ampere is looking to gain ground on Intel by building chips based on the Arm architecture, which is used in nearly every smartphone in the world, including Apple's iPhone and Samsung's Galaxy. The startup is trying to challenge Intel with better performance per watt and reduced prices.

Arm Holdings has also released chip designs dedicated to the the largest data centers, where high-end server chips can run thousands of dollars each. The company's new Neoverse chip designs use TSMC's most advanced production process, the 7-nanometer node. The chip designer has also collaborated in recent years with some of its customers including Huawei, Marvell, Ampere and Amazon to roll out server chips fabbed by TSMC.
 


Ric Ford

MacInTouch
For Apple, there's a lot of work involved, but the gains may be worth it. They wouldn't have to pay Intel for CPUs, and possibly not even AMD for GPUs, since all of the IP [intellectual property] would be available in-house. Those cost savings would mean higher margins for Apple.
I wonder, though, how this will affect the Intel-Apple relationship. I'm pretty sure Apple gets huge discounts on chips from Intel, considering the volumes of CPUs Apple sells across all its Macs. If Apple switches most of its Macs to ARMs, cutting out Intel, how is Intel going to respond – with prices, with delivery schedules, with new technology, with customization, etc.?

Oh, hey, here's a related drama...
Bloomberg said:
Ex-Apple Executive Accused of Betrayal Says He Was Snooped On
A former Apple Inc. executive who worked on the chips that power iPhones claims the company reviewed his confidential text messages before suing him for launching his own startup.

Gerard Williams III left Apple in February before starting Nuvia Inc. that same month with several other former Apple developers. Williams was the “chief architect” for all of Apple’s chips for its mobile devices. His new company is developing processors for use in data centers. In November, Nuvia exited stealth mode and announced funding worth $53 million.

Apple sued Williams for breach of contract in August, saying he was barred in an intellectual property agreement from planning or engaging in any business activities that are “competitive with or directly related to Apple’s business or products.”

In response, Williams accused Apple of a “stunning and disquieting invasion of privacy” over its monitoring of his texts. In one message, Williams said Apple would have “no choice but to purchase” his new company.
 


At the high end, you make a very relevant point, especially if having a macOS device doing it. But for consumer devices, it isn't that big of a stretch. Apple seems to show middling interest in the high-end workstation market, and as a product line, it makes me wonder if there is a future there. Porting is hard work, but how hard depends on the quality of the code you have.
It isn't all that hard to see iOS devices gain even more Mac-like capabilities and, in different configurations, become the future 'desktop'. But if there isn't a real future for pro desktops, yet removal of the line is of concern because of indirect impact on sales of other product lines, then doing an update every few years with very high cost makes sense. You don't have consumer-level marketing and stocking requirements. So the token effort that we seem to see looks like that sort of very long, very slow strangulation of a product line that is 'maturing out' of relevance.
But even if not: in order to do multiple-8K stream playback, you need a 28-core (56-thread) machine and an Afterburner card. You aren't doing that with a $10K Mac Pro. And that is relying on the Afterburner card to offload some of the work.

And throwing in extra ARM cores that are competitive with Intel's cores isn't an issue; 64-core ARM desktop machines are shipping now. Apple can customize to their heart's content, adding helpful instruction sets to speed the processing during the transition, and ship dual platforms (like when the laptops and desktops were on both PowerPC and x86).

The Mac Pro uses (probably) a 28-core Xeon W-3275M, which is a 200W processor that runs about $7500 retail for GeekBench of about 5000/60000 (taking the highest scores posted so far).

Now, an iPad Pro A12X single-core score is about the same at that Xeon's, at a mini-fraction of the power. A multi-core score for the Xeon of 60000 comes to about 2140 per core (slower clock speeds and all that), while the iPad Pro pulls about 18000 over its 8 cores to average about 2240 a core (and remember, that's 4 Lightning and 4 Thunder cores, optimized for mobile – eight Lightning cores would be even better) — at a vastly lower power draw. (18 W for the whole iPad system would be a reasonable estimate, with the display taking a large chunk of that number.)

So put 7 full iPads (not just CPUs) in the same cost line-item as the Xeon, and have 70% of the power draw (again, that's 7 of the full devices against just the Xeon CPU) and then run those, and it would be absolutely astonishing if 28 big cores plus the 28 little cores of ARM CPUs didn't compete with 28 cores on the Xeon.

Yes, there are workflows to make one platform look better than another. But the horsepower is already there. And, like Intel introduces special instructions (that are not universal across their CPUs), Apple could add similar extensions, but would have the ability to add the feature across the whole lineup, because they wouldn't need to hit a feature floor in order to produce chips for $120 laptops.

And then throw in the Afterburner card, since that is for offload, and it's easy to see the ARM machine potential; it should meet the performance while using significantly less energy to boot. Add in thermally constrained form factors (i.e. all the Macs outside of the Mac Pro), and there is even more headroom, since the same cooling apparatus would handle more horsepower on the ARM side.
 


Amazon disclaimer:
As an Amazon Associate I earn from qualifying purchases.

Latest posts