Mac Pro 2023 = Another Mac Pro 2013 crisis and failure.

[Longplays]

And the chart is misleading because it uses a quarter where Mac sales are unusually low relative to the rest of the company. Nowhere in your answer do I see an acknowledgement of that. When I get it wrong here on the site, I ackowledge it, and thank the poster for the correction. Why don't you?

With emphasis I was pointing out where the SoC R&D money is coming from. Macs would not have the revenue to self support their own SoC R&D with how unpopular they are relative to AMD/Intel.

Without the iPhone, Macs would still be stuck on 14nm Intel and Intel would still be at 14nm until now as no monopoly has any incentive to innovate.

Why are you so hung up on the the chart anyways? Does it say that Macs are as lucrative as iPhones? Is the variation that large to make it that wrong?

Or is my sample size too small for your standards?

I was thinking of using this chart below but visually I personally find it a bit busy, messy and unclean. Unlike a simple pie chart But then again the green of the "Mac Net Sales" is largely flat compared to other product/service lines.

When Gates was explaining the inapplicability of Moore's Law to cars, he was talking about year-over-year advances. That has nothing to do with my analogy, which is about the relative performance of the same tech in the same year used in different applications.

I think someone else answered your analogy succinctly.

Apple's chips are *so* efficient that, even with a quadratic or cubic increase in power with speed, they could implement a "turbo" boost for at least a few cores, giving significantly higher single-threaded performance in their desktop chips, while still having relatively low TDP. And that would be a big deal, since most apps today continue to be single-threaded. Hopefully they will do this with the M3.

I leave that design decision to Apple.

 

[Longplays]

You are rrying to prove one assumption with another assumption. This way one can prove anything.

Am I wrong when I say ~80% of all PCs shipped worldwide annually are laptops?

 

[VivienM]

Your premise is fundamentally wrong. 10 years ago and 30 years ago people were trying to use the computers for solving the problems they could do realistically with the compute power of the contemporary computers not the problems they wanted to solve. What you are saying is that Mac Pros are good enough to deal with compute tasks of the previous decade(s) and if one wants to do better they need to use modern computers (i.e. PCs)

No - what we're saying is that the number of tasks that need 'the absolute best performance money can buy' is shrinking. Ultimately, in most industries, what costs money is people, not computers. And, in particular, people waiting for computers to do tasks - so, in the world of 1992, if a $2000 Photoshop accelerator and a $6000 Quadra lets a graphic designer do twice as much Photoshop work in an 8 hour day, it's a slam dunk investment. You get effectively the benefit of two graphic designers for, assuming a two year amortization on that equipment, $4000/year - try hiring another person and giving them a Mac IIci for that much money!

The problem is that the number of workloads where this is true, i.e. where it makes sense to buy tens of thousands of dollars in equipment that will pay off in dramatically higher productivity, is shrinking. And as that number of tasks is shrinking, manufacturers are less and less likely to invest money in R&D to serve that shrinking market. And if anything, that means that as the high-end equipment's performance improves less quickly than the lower-end guys', this creates somewhat of a death spiral - plenty of businesses will spend $20K on equipment that makes a 4 hour task take 1 hour; no one will spend $20K on equipment that makes a 4 hour task take 3 hours and 55 minutes.

The x86 Windows guys' workstation business seems to be shrinking too. Intel has some advantage in that a lot of their workstation R&D can also be used for servers, at least for as long as people want to buy x64 servers, but...

Fundamentally, I'm not sure why you think x86 Windows guys are offering that much "modern" hardware that will massively outperform this new Apple Silicon Mac Pro. Are they? And more importantly, with the pace of innovation at TSMC, with Apple Silicon, etc, will that still be true in 5 years? It would not surprise me if your Apple Silicon MacBook Pro in 5 years could beat today's high-end x86 Windows workstations in all workloads.

 

[Longplays]

No - what we're saying is that the number of tasks that need 'the absolute best performance money can buy' is shrinking. Ultimately, in most industries, what costs money is people, not computers. And, in particular, people waiting for computers to do tasks - so, in the world of 1992, if a $2000 Photoshop accelerator and a $6000 Quadra lets a graphic designer do twice as much Photoshop work in an 8 hour day, it's a slam dunk investment. You get effectively the benefit of two graphic designers for, assuming a two year amortization on that equipment, $4000/year - try hiring another person and giving them a Mac IIci for that much money!

The problem is that the number of workloads where this is true, i.e. where it makes sense to buy tens of thousands of dollars in equipment that will pay off in dramatically higher productivity, is shrinking. And as that number of tasks is shrinking, manufacturers are less and less likely to invest money in R&D to serve that shrinking market. And if anything, that means that as the high-end equipment's performance improves less quickly than the lower-end guys', this creates somewhat of a death spiral - plenty of businesses will spend $20K on equipment that makes a 4 hour task take 1 hour; no one will spend $20K on equipment that makes a 4 hour task take 3 hours and 55 minutes.

The x86 Windows guys' workstation business seems to be shrinking too. Intel has some advantage in that a lot of their workstation R&D can also be used for servers, at least for as long as people want to buy x64 servers, but...

Fundamentally, I'm not sure why you think x86 Windows guys are offering that much "modern" hardware that will massively outperform this new Apple Silicon Mac Pro. Are they? And more importantly, with the pace of innovation at TSMC, with Apple Silicon, etc, will that still be true in 5 years? It would not surprise me if your Apple Silicon MacBook Pro in 5 years could beat today's high-end x86 Windows workstations in all workloads.

Looked at another way before 2008 the dominant smartphon brands/platforms were Motorola, BlackBerry, Palm, Nokia, etc.

Fast forward to 2023 and it is just Android + iPhone.

Android SoC and iPhone SoC are entering the old markets of laptops & desktops both consumer and workstation.

With their economies of scale and mastery of the ARM SoC of ~1.5 billion annually their R&D budget would outmuscle any x86 chip company.

Only thing in x86 corner is legacy hardware/software. Sounds eerily like a mainframe.

Before 2030s it will be interesting to see how ARM laptops will do in the PC space.

 

[falainber]

Am I wrong when I say ~80% of all PCs shipped worldwide annually are laptops?

That's not what you based your proof on. You referenced ARM laptop sales for Q4 2023 (which is a speculation).

 

[VivienM]

Before 2030s it will be interesting to see how ARM laptops will do in the PC space.

The other question is, why do you need a laptop at all? Smartphone docked USB-C to an external monitor with a keyboard/USB will probably be Good Enough for 90% of laptop workloads in 2030.

And it's worth noting - Apple has been putting all the pieces in place for that 'dockable smartphone' model over the last decade and a bit. I'm sure they would prefer to sell you a separate Mac, but if the industry did want dockable smartphones, their hardware and software stack is basically 98% there. Meanwhile, Microsoft, with their colossally catastrophic failure in smartphone/tablets, is... nowhere. And Android doesn't have the desktop side of the equation.

 

[falainber]

Looked at another way before 2008 the dominant mobile phone brands/platforms were Motorola, BlackBerry, Palm, Nokia, etc.

Fast forward to 2023 and it is just Android + iPhone.

Android SoC and iPhone SoC are entering the old markets of laptops & desktops both consumer and workstation.

With their economies of scale and mastery of the ARM SoC of ~1.5 billion annually their R&D budget would outcompete any x86 chip company.

Only thing in x86 corner is legacy hardware/software.

Before 2030s it will be interesting to see how ARM laptops will do in the PC space.

First Windows ARM based laptops were released in 2017. How much progress have they made in 6 years? None. Nobody is buying them. Apple ARM based laptops are irrelevant. Apple changed their processor architecture twice in the last couple of decades with marketshare of Macs staying basically unchanged.

 

[falainber]

The other question is, why do you need a laptop at all? Smartphone docked USB-C to an external monitor with a keyboard/USB will probably be Good Enough for 90% of laptop workloads in 2030.

And it's worth noting - Apple has been putting all the pieces in place for that 'dockable smartphone' model over the last decade and a bit. I'm sure they would prefer to sell you a separate Mac, but if the industry did want dockable smartphones, their hardware and software stack is basically 98% there. Meanwhile, Microsoft, with their colossally catastrophic failure in smartphone/tablets, is... nowhere. And Android doesn't have the desktop side of the equation.

In reality, Samsung phones (Dex) can already provide desktop experience when connected to the monitor. They are way ahead of Apple in this regard.

 

[Longplays]

First Windows ARM based laptops were released in 2017. How much progress have they made in 6 years? None. Nobody is buying them. Apple ARM based laptops are irrelevant. Apple changed their processor architecture twice in the last couple of decades with marketshare of Macs staying basically unchanged.

Someone pointed it out to me it may be better to go high-end ChromeOS instead of Windows considering porting x86 Win32 apps may not be of interest of PE companies.

Also Qualcomm's team are now ex-Apple engineers involved with Apple Silicon. That should improve things for ARM laptops.

With how many browser-based apps are out there... who knows what the next decade will bring.

Also check out this 10Y chart on desktop OS... macOS is nearing 19% worldwide while winOS is nearing 60%

 

[VivienM]

First Windows ARM based laptops were released in 2017. How much progress have they made in 6 years? None. Nobody is buying them. Apple ARM based laptops are irrelevant. Apple changed their processor architecture twice in the last couple of decades with marketshare of Macs staying basically unchanged.

Market share of Macs has changed big time from the peak of the dark era to today. With the biggest inflection point probably occuring around the beginning of the Intel era.

Walk into a post-secondary classroom in 2008 or, I presume, 2023, and you'd have seen tons and tons of MacBooks everywhere. 3-4 years earlier, it would have been a sea of Dells and HPs running XP.

Same thing in coffee shops.

I haven't observed it myself, but I am told that the market share of Macs in, say, software development is also way higher. And there are lots of young millennial companies that actually use Macs.

Now, I continue to think Windows on ARM is dumb. If your workloads are not tied to x64, why do you want Windows? But it's worth noting - no one has tried to make a high performance ARM chip for Windows. And part of the reason no one has is that the R&D in Windowsland is distributed. Microsoft doesn't design chips; Lenovo doesn't design chips; Intel designs chips but obviously isn't going to go and make a high end ARM chip for Windows. Folks like Qualcomm do design chips, but for lower-performing non-Windows devices and fundamentally, who is going to pay them in today's market to design a high-performance ARM SoC like Apple's Mx Pro/Max/Ultras for Windows use?

 

[VivienM]

In reality, Samsung phones (Dex) can already provide desktop experience when connected to the monitor. They are way ahead of Apple in this regard.

But... running what software?

Apple can easily give you the full desktop versions of software (in fact, give you access to the industry's second-biggest library of full-featured desktop software) when you dock your iPhone/iPad to an external monitor. Those full desktop versions already exist... they're called 'Mac apps'... and now they're recompiled for the same processor architecture powering your iPhone/iPad.

Neither Microsoft or Samsung, say, can give you mobile Word or Photoshop that exists in mobile form on your phone/tablet and magically morphs into full-featured desktop Word or Photoshop when you plug in your external monitor and peripherals. Apple has done essentially all the groundwork to be able to do so.

 

[Longplays]

Now, I continue to think Windows on ARM is dumb. If your workloads are not tied to x64, why do you want Windows? But it's worth noting - no one has tried to make a high performance ARM chip for Windows. And part of the reason no one has is that the R&D in Windowsland is distributed. Microsoft doesn't design chips; Lenovo doesn't design chips; Intel designs chips but obviously isn't going to go and make a high end ARM chip for Windows. Folks like

Employers with Windows PCs may want to continue with it given that the workforce are trained for Windows & Office.

Qualcomm do design chips, but for lower-performing non-Windows devices and fundamentally, who is going to pay them in today's market to design a high-performance ARM SoC like Apple's Mx Pro/Max/Ultras for Windows use?

Qualcomm bought NUVIA, a company founded by ex-Apple engineers that worked on Apple Silicon.

 

[VivienM]

Employers with Windows PCs may want to continue with it given that the workforce are trained for Windows & Office.

Or those employers are going to switch to running Windows/Office in a hosted VDI environment like Citrix, Windows 365, Amazon Workspaces, etc. If you have "legacy" apps and staff trained on Windows, and you're not doing resource-intensive things requiring bare metal on serious machines, those have been increasingly popular in the last decade or so.

And if you're running your Windows remotely, well, you don't need the client to be Windows.

 

[Longplays]

But... running what software?

Apple can easily give you the full desktop versions of software (in fact, give you access to the industry's second-biggest library of full-featured desktop software) when you dock your iPhone/iPad to an external monitor. Those full desktop versions already exist... they're called 'Mac apps'... and now they're recompiled for the same processor architecture powering your iPhone/iPad.

Neither Microsoft or Samsung, say, can give you mobile Word or Photoshop that exists in mobile form on your phone/tablet and magically morphs into full-featured desktop Word or Photoshop when you plug in your external monitor and peripherals. Apple has done essentially all the groundwork to be able to do so.

I'd be inclined to buy a 1TB or 2TB iPhone if I could run macOS when attached to an external display.

Hook it up via TB3 to a TB3 display that has I/O like TB3, USB-A 10Gbps, 1GbE, keyboard & mouse.

I'd be plenty happy with Core i7 performance even if it does not match M1 by a mile.

 

[Longplays]

Or those employers are going to switch to running Windows/Office in a hosted VDI environment like Citrix, Windows 365, Amazon Workspaces, etc. If you have "legacy" apps and staff trained on Windows, and you're not doing resource-intensive things requiring bare metal on serious machines, those have been increasingly popular in the last decade or so.

And if you're running your Windows remotely, well, you don't need the client to be Windows.

Think lower-end... more primitive workplace.

 

[VivienM]

I'd be inclined to buy a 1TB or 2TB iPhone if I could run macOS when attached to an external display.

Hook it up via TB3 to a TB3 display that has I/O like TB3, USB-A 10Gbps, 1GbE, keyboard & mouse.

I'd be plenty happy with Core i7 performance even if it does not match M1 by a mile.

And fundamentally, they have all of the pieces to offer you this today.

I think they will try their hardest NOT to actually ship that because they would rather sell you a Mac and an iPhone/iPad rather than get half the revenue from selling you a dockable iPhone/iPad.

But what's the old Steve Jobs line about how it's better to cannibalize yourself than to let someone else cannibalize you?

 

[Longplays]

And fundamentally, they have all of the pieces to offer you this today.

I think they will try their hardest NOT to actually ship that because they would rather sell you a Mac and an iPhone/iPad rather than get half the revenue from selling you a dockable iPhone/iPad.

But what's the old Steve Jobs line about how it's better to cannibalize yourself than to let someone else cannibalize you?

What do MR people say about "unupgradeable Macs"?.... e-waste! more e-waste!

I've been selling my iPhone after 2 years... no e-waste here!

 

[falainber]

Someone pointed it out to me it may be better to go high-end ChromeOS instead of Windows considering porting x86 Win32 apps may not be of interest of PE companies.

Also Qualcomm's team are now ex-Apple engineers involved with Apple Silicon. That should improve things for ARM laptops.

With how many browser-based apps are out there... who knows what the next decade will bring.

Also check out this 10Y chart on desktop OS... macOS is nearing 19% worldwide while winOS is nearing 60%

That's not a real market share. Firstly, StatCounter counts laptops as desktops. More importantly they just count the access to certain web sites. Apparently Mac users use their computers primarily for web browsing. Obviously, people who use computers for their work (mostly Windows) spend less time browsing internet. Statista puts Mac market share for desktops at 15% (https://www.statista.com/statistics/218089/global-market-share-of-windows-7/) but I suspect even this number is higher than the real Mac share simply because Mac sale numbers are very visible whereas PCs are sold by zillion of vendors, they get built by individuals and often used in ways that don't expose them on the internet.

 

[Longplays]

That's not a real market share. Firstly, StatCounter counts laptops as desktops. More importantly they just count the access to certain web sites. Apparently Mac users use their computers primarily for web browsing. Obviously, people who use computers for their work (mostly Windows) spend less time browsing internet. Statista puts Mac market share for desktops at 15% (https://www.statista.com/statistics/218089/global-market-share-of-windows-7/) but I suspect even this number is higher than the real Mac share simply because Mac sale numbers are very visible whereas PCs are sold by zillion of vendors, they get built by individuals and often used in ways that don't expose them on the internet.

Pls share higher quality data.

 

[falainber]

But... running what software?

Apple can easily give you the full desktop versions of software (in fact, give you access to the industry's second-biggest library of full-featured desktop software) when you dock your iPhone/iPad to an external monitor. Those full desktop versions already exist... they're called 'Mac apps'... and now they're recompiled for the same processor architecture powering your iPhone/iPad.

Neither Microsoft or Samsung, say, can give you mobile Word or Photoshop that exists in mobile form on your phone/tablet and magically morphs into full-featured desktop Word or Photoshop when you plug in your external monitor and peripherals. Apple has done essentially all the groundwork to be able to do so.

Actually, it's Apple that has very little to offer in terms of desktop apps. Is not FCP basically the only widely used desktop app from Apple? Regardless, it's only a speculation anyways whereas with Dex the list of Dex optimized apps already includes Microsoft Office suite, Adobe Acrobat Reader, Lightroom, Photoshop Sketch, Gmail, Chrome, BlueJeans, GoToMeeting and all the leading virtual desktop infrastructure (VDI) clients (and more).

 

[falainber]

Pls share higher quality data.

I already did. Statista does provide higher quality data than StatCounter.

 

[bobcomer]

Apple's chips are *so* efficient that, even with a quadratic or cubic increase in power with speed, they could implement a "turbo" boost for at least a few cores, giving significantly higher single-threaded performance in their desktop chips, while still having relatively low TDP.

I really like that idea.

 

[VivienM]

Actually, it's Apple that has very little to offer in terms of desktop apps. Is not FCP basically the only widely used desktop app from Apple? Regardless, it's only a speculation anyways whereas with Dex the list of Dex optimized apps already includes Microsoft Office suite, Adobe Acrobat Reader, Lightroom, Photoshop Sketch, Gmail, Chrome, BlueJeans, GoToMeeting and all the leading virtual desktop infrastructure (VDI) clients (and more).

I'm counting all the third party Apple Silicon-native apps for "Mac" because, well, if Apple makes a dockable iPhone/iPad, they'll run on it.

Mac, at least Intel, has the second-largest library of native desktop apps. And those apps are being recompiled to run on something that is equivalent to iPhone/iPads in all material ways.

 

[bobcomer]

Think lower-end... more primitive workplace.

That's a good point, and where I'm at. We just don't have the $'s necessary to get an internet connection that could handle all our workers using VDI and basically dumb terminals, and the internet just isn't that reliable enough anyway -- my IT spidey sense is yelling in alarm at the thought. Due to what we make, we have to have a lot of water, so by a river in a non populated region = poor internet = No VDI unless it's locally served and that is actually more expensive to do than consumer PC hardware.

 

[Longplays]

That's a good point, and where I'm at. We just don't have the $'s necessary to get an internet connection that could handle all our workers using VDI and basically dumb terminals, and the internet just isn't that reliable enough anyway -- my IT spidey sense is yelling in alarm at the thought. Due to what we make, we have to have a lot of water, so by a river in a non populated region = poor internet = No VDI unless it's locally served and that is actually more expensive to do than consumer PC hardware.

I look forward to the day when x86 will just be legacy. Like mainframes.