Right, so if Parallels has been recompiled for Apple Silicon, couldn't it then run Windows in a VM?
Parallels would run, yes, and it could run virtual machines running operating systems compiled to run on ARM as well as any software that will compile, but it will not run x86-based operating systems such as Windows (or Linux compiled for x86).
Ah. I guess I thought the virtualization software could emulate the underlying processor needed as well.
There is software like qemu that can emulate almost any CPU but it is slow. It might be fine for running Windows 95 or NT but anything from Windows 7 up is going to be painful unless the Apple Silicon CPUs are insanely fast. I expect great things from Apple's efforts but there are limits.
On the other hand, apparently Apple's Rosetta 2 does a fine job trans-compiling (transpiling) from x86_64 to AArch64 (ARM instructions). If Apple would open Rosetta 2 up for VM developers then it might be possible to get decent speed from translated x86 binary code instead of emulation. But right now, it looks like Apple is either uninterested in doing this or for some technical reason finds it not feasible. I suspect part of the problem is that Rosetta 2 doesn't translate anything but 64-bit x86_64 and perhaps Windows does not restrict itself sufficiently to meet that requirement. Apple has been planning this transition for a while and may have made sure that their developer tools only produced explicit x86_64 code. If someone with more expertise could weigh in, I would love to hear if this is possibly the problem.
The Mac's best days are behind it now. And I'm not coming along for the ride, I'm done. I'll still work on the ARM Macs for clients, but I will never own one. And I'm going to be brushing up on my Linux and Windoze support skills, looks like I'm going to need it.
Apple just killed Macintosh.
HA! I've read this SO MANY TIMES over the past 30 YEARS there's no way I can't just laugh now. DOOMED.
It's sad that Apple is in effect putting Parallels out of business.
Yeah except for the Apple Silicon compile of Parallels they literally demoed Monday.
They've certainly limited their market. I can't imagine Parallels userbase skews towards people who run ARM-compiled Linux distros.
You know, we could just wait a minute and see what actually ends up happening. Notice they didn't bother to edit out the Windows logo out of the Parallels icon? Not saying that's a thing, but let's not jump to conclusions.
-- With an A series processor a MacBook will essentially be an IPad without a touchscreen but an arguably better OS.
What tf are you smoking? It's a Mac, running macOS but the only difference now is it's on a different processor.
What am I smoking? It's called "reality".
And that reality is: it will be an enhanced iPad in a clamshell design with a larger screen (but without, we assume, a touchscreen) and with, as I said, an arguably better OS.
This is a mistake. This is a move away from professional work to “grandma wordprocessing”. Interoperability. Unix base. Bootcamp. They did a lot to undermine Mac’s reputation that it is a toy computer and here Apple goes again trying to destroy everything.
If you work in development, AML, cybersec, pentesting, or distributed networking, virtualization isn’t recommended—it’s required. Sharing a 64 bit cpu co-operability with Windows, Linux and BSD means that virtualization works natively and out of the box. Having to run resource limited code through an emulating interpreter is a devastating nonstarter.
Forget gaming. Gaming is already a wasteland on Mac but it’ll get worse after that. Forget large, complex, robust titles like GTA VI, Diablo IV or the like. If coding wrappers for Open GL to DX12 was a pain before, and moving to only 64bit code wrecked games as devs refused to expend resources to fix the issue. Anyone that thinks that Apple has gaming on iOS is lying to themselves. Most titles are ten year old remakes or microtransactional vomit stew. And no, the occasional indie title doesn’t make up for saying goodbye to companies like iD, Blizzard, Crystal Dynamics, et al.
If my neighbors computer can do work, play games and inter operate out of the box, while Macs need Wintel machines or consoles on top of of the already purchase price to get stuff done, you’ve diminished the platforms utility.
This isn’t about speed. Apple A series chips may get close to Intels Coffee Lake i5/7 levels of performance but they’ll lag compared to Golden Coves 10 nm cores next year. If Apple was concerned about security or supply chain, they could’ve gone AMD. Ryzen 3 4xxx series chips have a 15% uplift over 3700x or Core i7s in the 50W envelope and they have a great relationship with AMD. No one will take enterprise seriously with Mac not because A series chips are weak, although they surely are against EPYC, Xeon, Intel HEDT, Vega II, Ampere... etc but because no one will rewrite entire code based for Apple.
Apple software development will be based solely on scraps left over from iOS. No enterprise. No AML. No shared development. No real gaming. No virtualization. Macs become 3000 dollar email terminals. This is devastating news for the Mac platform.
This is a mistake. This is a move away from professional work to “grandma wordprocessing”. Interoperability. Unix base. Bootcamp. They did a lot to undermine Mac’s reputation that it is a toy computer and here Apple goes again trying to destroy everything.
If you work in development, AML, cybersec, pentesting, or distributed networking, virtualization isn’t recommended—it’s required. Sharing a 64 bit cpu co-operability with Windows, Linux and BSD means that virtualization works natively and out of the box. Having to run resource limited code through an emulating interpreter is a devastating nonstarter.
Forget gaming. Gaming is already a wasteland on Mac but it’ll get worse after that. Forget large, complex, robust titles like GTA VI, Diablo IV or the like. If coding wrappers for Open GL to DX12 was a pain before, and moving to only 64bit code wrecked games as devs refused to expend resources to fix the issue. Anyone that thinks that Apple has gaming on iOS is lying to themselves. Most titles are ten year old remakes or microtransactional vomit stew. And no, the occasional indie title doesn’t make up for saying goodbye to companies like iD, Blizzard, Crystal Dynamics, et al.
If my neighbors computer can do work, play games and inter operate out of the box, while Macs need Wintel machines or consoles on top of of the already purchase price to get stuff done, you’ve diminished the platforms utility.
This isn’t about speed. Apple A series chips may get close to Intels Coffee Lake i5/7 levels of performance but they’ll lag compared to Golden Coves 10 nm cores next year. If Apple was concerned about security or supply chain, they could’ve gone AMD. Ryzen 3 4xxx series chips have a 15% uplift over 3700x or Core i7s in the 50W envelope and they have a great relationship with AMD. No one will take enterprise seriously with Mac not because A series chips are weak, although they surely are against EPYC, Xeon, Intel HEDT, Vega II, Ampere... etc but because no one will rewrite entire code based for Apple.
Apple software development will be based solely on scraps left over from iOS. No enterprise. No AML. No shared development. No real gaming. No virtualization. Macs become 3000 dollar email terminals. This is devastating news for the Mac platform.
Let's hope that you're wrong.
Apple is not infallible. And, one of their greatest strengths is that they know that. They know it very, very well. So, I suspect (and hope) that they thought of your concerns / predictions and have a plan.
I find all of the doom and gloom predictions of the demise of Apple because they are appropriately abandoning the x86 platform for very legitimate reasons quite comical.
I myself remain tied to the PowerPC platform and still using a G5 which is considered "ancient" by modern standards. But kudos for the tenfourfox developers as my G5 remains relevant. More than I can say for my 2008 iMac using a core2 duo CPU.
It was actually a sad day for me when Apple abandoned the PowerPC especially when PA Semi was manufacturing the Pwrficient CPU which offered great performance at low power consumption figures. Something that Intel has still failed to come to grips with. Admittedly Intel's relationship with Apple has helped. But Intel is totally lost. AMD is outperforming Intel while Intel continued to flounder. Still on a 14 nm process when everyone else has moved beyond. And Intel's integrated GPU is still an atrocious performer.
Intel is stuck and their efforts are anemic at best. Improvements to their CPU are occurring at even slower than a snail's pace.
Suddenly those toasted bunnies from that Apple commercial all those years ago seems eerily prescient.
So for all of those who yearn for the x86 CPU, just buy a modern Mac today. The CPU will still be relevant a decade from now. Because Intel will still have not have advanced much even by then. Today's Intel CPUs will not be obsoleted by newer Intel CPUs for a long time. If ever.
Apple can either stay tied to Intel and continue to advance the A series CPUs for their mobile devices and watch the next generation iPad move past the MacBook Pro even with all of the limitations of the mobile platform or bite the bullet and develop a high performance chip based on the ARM ISA for the MacBook Pro and even the Mac Pro line of machines.
I mean look at the optics people. The iPad pro is matching the core i7 in single threaded performance and in a mobile device!!! Even with much lower amounts of RAM, passive cooling and a limited thermal envelope, Apples bionic CPU is on par with the i7!! For Intel, that's actually not just pathetic, it's embarrassing.
Why wouldn't Apple just ditch Intel. Apple clearly has a much better CPU with a superior integrated GPU than Intel. Why not just develop a high performance version for desktop and laptop use? Honestly, an Intel CPU will remain relevant for a very long time. If nothing more because Intel isn't able to advance the x86 platform. It's even more pathetic that AMD offers better performance these days.
For those that require an x86 CPU for Windows/DOS compatibility, well they will be fine for a long time. Windows will be tied to the x86 platform for a long time to come. And Intel isn't exactly moving forward very fast.
It makes no sense for Apple to advance their mobile platforms while the desktop and laptop lines languish because of Intel's ineptitude.
Moving to a high performance Apple CPU based on the ARM ISA makes complete sense. And the move will remove Intel's management engine entirely. It is a security nightmare.
To continue with my previous post, moving to AMD with it's secure performance autonomous subsystem is just as bad as Intel's management engine. Apple's own CPU eliminates all of that mess. It's the primary reason I use a G5 for my desktop needs.
Admittedly, I don't make a living where high performance processing is required. So the decision to use the PowerPC for security reasons works for me.
So this all boils down to those who say they need the highest performing CPUs but it has to be an x86 platform makes no sense. Because the Apple CPU IS going to trounce anything from Intel. And if running legacy applications is necessary, then high performance processing isn't exactly necessary.
And if someone has 50k to pay for a high end multi CPU Mac Pro, then they will be willing to spend less on a higher performing Mac Pro with Apple CPU, even if it requires a dongle with an x86 CPU attached to a Thunderbolt/USB4 port to run legacy programs. Because all of the programs that require all of that CPU power will be ported over anyhow. And legacy programs requiring Windows/DOS will be fine on any modern x86 cpu for a very long time.
I just don't buy the argument that high end games won't make it over to the Apple CPU based Mac. Because high performance CPU/GPU combinations have always been gamers' wet dreams. And the Apple CPU/SOC will offer such in spades.
What will the Windows apologists say when a low end Apple platforms out performs a high performance and much more costly windows one?
Even if Apple gets special pricing on Intel products, there is no way Intel can match the price of Apple's own chips. Apple already spends a small fortune developing those ARM based processors for their mobile products. They might as well extend that expertise in developing high performance desktop SOCs. Most of the development cost is already baked in anyway.
Apple will totally own the security of their own platform and not worry about the management engine. They can much more easily customize the silicon for the needs of their user base.
Apple can much more rapidly develop their own platform. And the rest of the industry will be playing catch up. Even Microsoft will see the urgency of moving Windows over to the ARM platform. Virtualizing the ARM version of Windows for the Mac platform will be the ONLY way for Microsoft to remain relevant for the long term.
The majority of software development is taking place for ARM. Not x86. Not by a long shot.
Apple is going to outflank both Microsoft and Google/Alphabet with one decisive movement in taking this step.
The only sad thing is that current world events may make the whole matter moot. I mean, when the economy implodes, this whole thing becomes completely irrelevant.
I got a dual boot thinkpad with Linux and Windows, just in case!
The potential issues for me:
Windows apps I must have. These are becoming very few: MS Office, a Windows only music scanning program, etc. I could actually dump Windows if I had to. Not Apple Silicon Ready
Music production: You can do it on Linux, you can sort of do it in Windows, but Macs rule this area. Sure Bitwig will run on any OS, but I like Logic. I can start a project in GarageBand on iOS, and then work on it in Logic. - Apple Silicon Ready
DevOps/Web Development: for me at the moment, this means docker. I can do the same work on Mac, Linux or Windows (in WSL2 Linux). Apple Silicon Ready
OpenGL dependent apps: Many of these will work or will be ported, probably with some bumps in the road. They already started running worse as of Maverick. With a complete switch to Metal and Apple Silicon, I may have to switch to the Linux box until they are ported. Some of that will not be trivial. This includes gimp, Inkscape, blender and viewers for open simulator. Not Future MAC Ready....yet?
VR/3D apps: I was already to run out and get a sonnet puck external GPU as it works with Mac or Wintel. I was interested in the oculus rift as well. Perhaps for the Lintel/Wintel box. Apple is rumored to be doing something in the VR space, but there are no hints yet. Not Apple Silicon Ready for me
Virtual machines: only as long as it is ARM. This will eventually be functional, but it is too much of a break in my existing stuff for now. Not Apple Silicon Ready - for my workflow.
My most important commercial apps are on the Mac, and some only on the Mac, but I also want all of the great linux apps I use. My conclusion is that any performance wins from Apple Silicon are not enough to offset the disruption to my tools and workflow in the near future. My next Mac will still be an x86 box.
BUT I can see the potential long term gains, and I think that Apple is making the right move, and not all of it is even purely self-serving. Depending on how it all shakes out, and how long non-trivial stuff takes to port, I could see the Mac after next being Apple Silicon.
I find all of the doom and gloom predictions of the demise of Apple because they are appropriately abandoning the x86 platform for very legitimate reasons quite comical.
I myself remain tied to the PowerPC platform and still using a G5 which is considered "ancient" by modern standards. But kudos for the tenfourfox developers as my G5 remains relevant. More than I can say for my 2008 iMac using a core2 duo CPU.
Talking about real work, while rocking 12 year old GPULs that have less performance than a Core i5 3650 is quite comical.
It was actually a sad day for me when Apple abandoned the PowerPC especially when PA Semi was manufacturing the Pwrficient CPU which offered great performance at low power consumption figures. Something that Intel has still failed to come to grips with. Admittedly Intel's relationship with Apple has helped. But Intel is totally lost. AMD is outperforming Intel while Intel continued to flounder. Still on a 14 nm process when everyone else has moved beyond. And Intel's integrated GPU is still an atrocious performer.
Ryzen 2 is still behind Intel in AVX/decode. Ryzen 2 is still behind Intel in single core and single thread performance. Yes AMD IS VERY COMPETITIVE IN CPUs right now but let’s not pretend that Intel Rocket or Cannon lakes are the same as oh... I don’t know, an Apple GPUL chip. That would be comical.
Intel is stuck and their efforts are anemic at best. Improvements to their CPU are occurring at even slower than a snail's pace.
Actually, their upcoming Golden Cove cores should compete well against Ryzen 3. They also finally smoothed out their 10 nm lithography which is competitive with TSMC 7nm.
Suddenly those toasted bunnies from that Apple commercial all those years ago seems eerily prescient.
You’re exaggerating Intel’s ills. Pretty comical. I’m not an Intel fan but they still supply great desktop chips with elite performance. They also have virtualized interoperability throughout the technical world.
So for all of those who yearn for the x86 CPU, just buy a modern Mac today. The CPU will still be relevant a decade from now. Because Intel will still have not have advanced much even by then. Today's Intel CPUs will not be obsoleted by newer Intel CPUs for a long time. If ever.
Not only comical but also erroneous. Per your argument a 10 year old 2 core Penryn does the same as a Core i9 10800x. Except the latter is actually 1200% faster than the former In multithreaded. In ten years that’s a good uplift by any standards, even Apples.
Apple can either stay tied to Intel and continue to advance the A series CPUs for their mobile devices and watch the next generation iPad move past the MacBook Pro even with all of the limitations of the mobile platform or bite the bullet and develop a high performance chip based on the ARM ISA for the MacBook Pro and even the Mac Pro line of machines.
I mean look at the optics people. The iPad pro is matching the core i7 in single threaded performance and in a mobile device!!! Even with much lower amounts of RAM, passive cooling and a limited thermal envelope, Apples bionic CPU is on par with the i7!! For Intel, that's actually not just pathetic, it's embarrassing.
Except it’s not matching a modern i7. The A12Z matches up with a Core i5 Mobility 10000 series. Sure, they could conceivably add 4 more big cores and increase cooling, but theyd only probably hit Coffee Lake desktop i7 speeds. To catch Intels best from the new i9s, theyd have to increase performance over the A12Z by 100%. Can it be done? Sure. It’s still an ARM chip and while Apple will integrate the software well through optimizations, you’d have to be delusional to think that 3rd party devs would do the same. The fact is You’re going to be eating a performance hit for the next five years or so regardless of what Apple does with their chips.
Why wouldn't Apple just ditch Intel. Apple clearly has a much better CPU with a superior integrated GPU than Intel. Why not just develop a high performance version for desktop and laptop use? Honestly, an Intel CPU will remain relevant for a very long time. If nothing more because Intel isn't able to advance the x86 platform. It's even more pathetic that AMD offers better ...
I skipped the rest for redundancy. Do you know what will make Apple computers less safe? The fact that you can’t run other virtualized oses for penetration testing or cyber security testing. What this means is that hackers that focus on Apple computers will have almost free reign since Apple will have little to no similarities with other platforms and won’t be able to leverage the knowledge, expertise and participation of members of other platforms.
-- With an A series processor a MacBook will essentially be an IPad without a touchscreen but an arguably better OS.
What tf are you smoking? It's a Mac, running macOS but the only difference now is it's on a different processor.
What am I smoking? It's called "reality".
And that reality is: it will be an enhanced iPad in a clamshell design with a larger screen (but without, we assume, a touchscreen) and with, as I said, an arguably better OS.
Reality is it's a Mac running macOS, full stop. The processor is irrelevant.
This is a mistake. This is a move away from professional work to “grandma wordprocessing”. Interoperability. Unix base. Bootcamp. They did a lot to undermine Mac’s reputation that it is a toy computer and here Apple goes again trying to destroy everything.
If you work in development, AML, cybersec, pentesting, or distributed networking, virtualization isn’t recommended—it’s required. Sharing a 64 bit cpu co-operability with Windows, Linux and BSD means that virtualization works natively and out of the box. Having to run resource limited code through an emulating interpreter is a devastating nonstarter.
Forget gaming. Gaming is already a wasteland on Mac but it’ll get worse after that. Forget large, complex, robust titles like GTA VI, Diablo IV or the like. If coding wrappers for Open GL to DX12 was a pain before, and moving to only 64bit code wrecked games as devs refused to expend resources to fix the issue. Anyone that thinks that Apple has gaming on iOS is lying to themselves. Most titles are ten year old remakes or microtransactional vomit stew. And no, the occasional indie title doesn’t make up for saying goodbye to companies like iD, Blizzard, Crystal Dynamics, et al.
If my neighbors computer can do work, play games and inter operate out of the box, while Macs need Wintel machines or consoles on top of of the already purchase price to get stuff done, you’ve diminished the platforms utility.
This isn’t about speed. Apple A series chips may get close to Intels Coffee Lake i5/7 levels of performance but they’ll lag compared to Golden Coves 10 nm cores next year. If Apple was concerned about security or supply chain, they could’ve gone AMD. Ryzen 3 4xxx series chips have a 15% uplift over 3700x or Core i7s in the 50W envelope and they have a great relationship with AMD. No one will take enterprise seriously with Mac not because A series chips are weak, although they surely are against EPYC, Xeon, Intel HEDT, Vega II, Ampere... etc but because no one will rewrite entire code based for Apple.
Apple software development will be based solely on scraps left over from iOS. No enterprise. No AML. No shared development. No real gaming. No virtualization. Macs become 3000 dollar email terminals. This is devastating news for the Mac platform.
-- With an A series processor a MacBook will essentially be an IPad without a touchscreen but an arguably better OS.
What tf are you smoking? It's a Mac, running macOS but the only difference now is it's on a different processor.
What am I smoking? It's called "reality".
And that reality is: it will be an enhanced iPad in a clamshell design with a larger screen (but without, we assume, a touchscreen) and with, as I said, an arguably better OS.
Reality is it's a Mac running macOS, full stop. The processor is irrelevant.
Interestingly, that's what I've been saying all along: What sets Macs (and iPhones) apart is not the hardware but the software and Apple's ecosystem.
But, with every new release, the pundits and experts feverishly compare hardware to hardware trying to figure which has the edge -- and we follow right along with it. And, admittedly, I fell into the same trap: when I said the MacBook will be a glorified iPad with a clamshell design I was speaking strictly from the hardware level.
... But too, I should have also qualified that by saying: with a bigger shelll, less emphasis on weight and even better cooling available, Apple could enhance it to surpass the iPad.
However, since Apple's instruction set is no longer that similar to ARM, I think that's unfortunately unlikely.
Seems you haven't been paying attention for the past ten years or so.
Apple's silicon uses the ARM instruction set.
Barely. It's changed so much since the the 5S days when it was announced that Apple was using AArch64. Apple has zero need for backwards compatibility since object code is uploaded to the App Store from Xcode rather than byte code. The App Store then compiles the correct version for each device, doesn't matter what the architecture is then.
Have you got any evidence for this? It's unlikely to be correct, judging by how fast they've got Docker and Linux running. While I suspect they've added extensions to the instruction set to optimise their own devices, creating their own instruction set from scratch … I don't see it myself.
Read an interesting thread on ArsTechnica, which seems to mirror a lot of the conversations here.
Someone pointed out that only 2% of the Macs use Bootcamp.
Someone else pointed out that there are other ways to run Windows without Bootcamp.
But, as person number 1 pointed out, these days, many virtualisation setups use Bootcamp to power the Windows install, so the number of users who actually need to run Windows on a Mac is not that far from 2%.
As a .Net developer who's used Macs with ParalIels or VMWare for over 10 years I pooped my pants when I heard this news, so, I decided yesterday to go Mac native and transitioned over to the Mac version of Visual Studio and installed SQL server in a Docker container on the Mac, Azure data Studio and SQL Pro for MMSQL replace SSMS.
VS Mac has a couple of really insignificant features missing such as adding code bookmarks and the ability to easily tweak the colors of the IDE (I did say insignificant but I use these and you only realize how much when they're not there!) but so far am impressed with the build speed using Visual Studio for Mac when compared to same project yesterday using Parallels with VS2019 on Windows. A Blazor project that I'm working on consistently compiles in under 10 seconds compared to 25-30.
The Mac's best days are behind it now. And I'm not coming along for the ride, I'm done. I'll still work on the ARM Macs for clients, but I will never own one. And I'm going to be brushing up on my Linux and Windoze support skills, looks like I'm going to need it.
Apple just killed Macintosh.
HA! I've read this SO MANY TIMES over the past 30 YEARS there's no way I can't just laugh now. DOOMED.
It's wishful thinking basically. Whenever Apple does something like this, some people just take it personally. And when you read back on some of the positions they took earlier on, I think the problem is they feel Apple has made them look a bit foolish, especially those who claim to know Apple inside out.
Apple is out of excuses not to switch to AMD now. And no Apple isn't switching to ARM and yes they have been testing Zen in-house for over a year.
And looking at this:
They will come out before Apple comes out with their own chips which will be ... never. Seriously, I wish people will give this up. It isn't their area of expertise. CPU and SOC design isn't easy. If it was, everyone would do it. Instead there are only two CPU companies on the planet and have been for decades - Intel and AMD both of whom use the same base x86 design - and the number of SOC companies isn't that much bigger (and again they all start from the same base ARM Holdings design). Sun Microsystems, Motorola and IBM, who were all making CPUs or SOCs as recently as the 1990s? RIP. Also there is this whole "application compatibility" thing. Macs run on x86 just like Windows and Linux computers. Result: while you have to tweak for the different OSes and such, all "PC" applications are developed for the x86 instruction set. If Apple wants to use their ARM SOCs for MacBooks or develop a wholly new SOC/CPU, all those developers would have to port, rewrite or create from scratch their x86 applications or Apple would have to emulate x86 (more on this later).
Perhaps they think developers still write in machine code?
Comments
On the other hand, apparently Apple's Rosetta 2 does a fine job trans-compiling (transpiling) from x86_64 to AArch64 (ARM instructions). If Apple would open Rosetta 2 up for VM developers then it might be possible to get decent speed from translated x86 binary code instead of emulation. But right now, it looks like Apple is either uninterested in doing this or for some technical reason finds it not feasible. I suspect part of the problem is that Rosetta 2 doesn't translate anything but 64-bit x86_64 and perhaps Windows does not restrict itself sufficiently to meet that requirement. Apple has been planning this transition for a while and may have made sure that their developer tools only produced explicit x86_64 code. If someone with more expertise could weigh in, I would love to hear if this is possibly the problem.
What tf are you smoking? It's a Mac, running macOS but the only difference now is it's on a different processor.
Yeah except for the Apple Silicon compile of Parallels they literally demoed Monday.
If you work in development, AML, cybersec, pentesting, or distributed networking, virtualization isn’t recommended—it’s required. Sharing a 64 bit cpu co-operability with Windows, Linux and BSD means that virtualization works natively and out of the box. Having to run resource limited code through an emulating interpreter is a devastating nonstarter.
Forget gaming. Gaming is already a wasteland on Mac but it’ll get worse after that. Forget large, complex, robust titles like GTA VI, Diablo IV or the like. If coding wrappers for Open GL to DX12 was a pain before, and moving to only 64bit code wrecked games as devs refused to expend resources to fix the issue. Anyone that thinks that Apple has gaming on iOS is lying to themselves. Most titles are ten year old remakes or microtransactional vomit stew. And no, the occasional indie title doesn’t make up for saying goodbye to companies like iD, Blizzard, Crystal Dynamics, et al.
This isn’t about speed. Apple A series chips may get close to Intels Coffee Lake i5/7 levels of performance but they’ll lag compared to Golden Coves 10 nm cores next year. If Apple was concerned about security or supply chain, they could’ve gone AMD. Ryzen 3 4xxx series chips have a 15% uplift over 3700x or Core i7s in the 50W envelope and they have a great relationship with AMD. No one will take enterprise seriously with Mac not because A series chips are weak, although they surely are against EPYC, Xeon, Intel HEDT, Vega II, Ampere... etc but because no one will rewrite entire code based for Apple.
Apple software development will be based solely on scraps left over from iOS. No enterprise. No AML. No shared development. No real gaming. No virtualization. Macs become 3000 dollar email terminals. This is devastating news for the Mac platform.
I myself remain tied to the PowerPC platform and still using a G5 which is considered "ancient" by modern standards. But kudos for the tenfourfox developers as my G5 remains relevant. More than I can say for my 2008 iMac using a core2 duo CPU.
It was actually a sad day for me when Apple abandoned the PowerPC especially when PA Semi was manufacturing the Pwrficient CPU which offered great performance at low power consumption figures. Something that Intel has still failed to come to grips with. Admittedly Intel's relationship with Apple has helped. But Intel is totally lost. AMD is outperforming Intel while Intel continued to flounder. Still on a 14 nm process when everyone else has moved beyond. And Intel's integrated GPU is still an atrocious performer.
Intel is stuck and their efforts are anemic at best. Improvements to their CPU are occurring at even slower than a snail's pace.
Suddenly those toasted bunnies from that Apple commercial all those years ago seems eerily prescient.
So for all of those who yearn for the x86 CPU, just buy a modern Mac today. The CPU will still be relevant a decade from now. Because Intel will still have not have advanced much even by then. Today's Intel CPUs will not be obsoleted by newer Intel CPUs for a long time. If ever.
Apple can either stay tied to Intel and continue to advance the A series CPUs for their mobile devices and watch the next generation iPad move past the MacBook Pro even with all of the limitations of the mobile platform or bite the bullet and develop a high performance chip based on the ARM ISA for the MacBook Pro and even the Mac Pro line of machines.
I mean look at the optics people. The iPad pro is matching the core i7 in single threaded performance and in a mobile device!!! Even with much lower amounts of RAM, passive cooling and a limited thermal envelope, Apples bionic CPU is on par with the i7!! For Intel, that's actually not just pathetic, it's embarrassing.
Why wouldn't Apple just ditch Intel. Apple clearly has a much better CPU with a superior integrated GPU than Intel. Why not just develop a high performance version for desktop and laptop use? Honestly, an Intel CPU will remain relevant for a very long time. If nothing more because Intel isn't able to advance the x86 platform. It's even more pathetic that AMD offers better performance these days.
For those that require an x86 CPU for Windows/DOS compatibility, well they will be fine for a long time. Windows will be tied to the x86 platform for a long time to come. And Intel isn't exactly moving forward very fast.
It makes no sense for Apple to advance their mobile platforms while the desktop and laptop lines languish because of Intel's ineptitude.
Moving to a high performance Apple CPU based on the ARM ISA makes complete sense. And the move will remove Intel's management engine entirely. It is a security nightmare.
Admittedly, I don't make a living where high performance processing is required. So the decision to use the PowerPC for security reasons works for me.
So this all boils down to those who say they need the highest performing CPUs but it has to be an x86 platform makes no sense. Because the Apple CPU IS going to trounce anything from Intel. And if running legacy applications is necessary, then high performance processing isn't exactly necessary.
And if someone has 50k to pay for a high end multi CPU Mac Pro, then they will be willing to spend less on a higher performing Mac Pro with Apple CPU, even if it requires a dongle with an x86 CPU attached to a Thunderbolt/USB4 port to run legacy programs. Because all of the programs that require all of that CPU power will be ported over anyhow. And legacy programs requiring Windows/DOS will be fine on any modern x86 cpu for a very long time.
I just don't buy the argument that high end games won't make it over to the Apple CPU based Mac. Because high performance CPU/GPU combinations have always been gamers' wet dreams. And the Apple CPU/SOC will offer such in spades.
What will the Windows apologists say when a low end Apple platforms out performs a high performance and much more costly windows one?
Even if Apple gets special pricing on Intel products, there is no way Intel can match the price of Apple's own chips. Apple already spends a small fortune developing those ARM based processors for their mobile products. They might as well extend that expertise in developing high performance desktop SOCs. Most of the development cost is already baked in anyway.
Apple will totally own the security of their own platform and not worry about the management engine. They can much more easily customize the silicon for the needs of their user base.
Apple can much more rapidly develop their own platform. And the rest of the industry will be playing catch up. Even Microsoft will see the urgency of moving Windows over to the ARM platform. Virtualizing the ARM version of Windows for the Mac platform will be the ONLY way for Microsoft to remain relevant for the long term.
The majority of software development is taking place for ARM. Not x86. Not by a long shot.
Apple is going to outflank both Microsoft and Google/Alphabet with one decisive movement in taking this step.
The only sad thing is that current world events may make the whole matter moot. I mean, when the economy implodes, this whole thing becomes completely irrelevant.
The potential issues for me:
Windows apps I must have. These are becoming very few: MS Office, a Windows only music scanning program, etc. I could actually dump Windows if I had to. Not Apple Silicon Ready
Music production: You can do it on Linux, you can sort of do it in Windows, but Macs rule this area. Sure Bitwig will run on any OS, but I like Logic. I can start a project in GarageBand on iOS, and then work on it in Logic. - Apple Silicon Ready
DevOps/Web Development: for me at the moment, this means docker. I can do the same work on Mac, Linux or Windows (in WSL2 Linux). Apple Silicon Ready
OpenGL dependent apps: Many of these will work or will be ported, probably with some bumps in the road. They already started running worse as of Maverick. With a complete switch to Metal and Apple Silicon, I may have to switch to the Linux box until they are ported. Some of that will not be trivial. This includes gimp, Inkscape, blender and viewers for open simulator. Not Future MAC Ready....yet?
VR/3D apps: I was already to run out and get a sonnet puck external GPU as it works with Mac or Wintel. I was interested in the oculus rift as well. Perhaps for the Lintel/Wintel box. Apple is rumored to be doing something in the VR space, but there are no hints yet. Not Apple Silicon Ready for me
Virtual machines: only as long as it is ARM. This will eventually be functional, but it is too much of a break in my existing stuff for now. Not Apple Silicon Ready - for my workflow.
My most important commercial apps are on the Mac, and some only on the Mac, but I also want all of the great linux apps I use. My conclusion is that any performance wins from Apple Silicon are not enough to offset the disruption to my tools and workflow in the near future. My next Mac will still be an x86 box.
BUT I can see the potential long term gains, and I think that Apple is making the right move, and not all of it is even purely self-serving. Depending on how it all shakes out, and how long non-trivial stuff takes to port, I could see the Mac after next being Apple Silicon.
So you're one of those glass half-full people.
Someone pointed out that only 2% of the Macs use Bootcamp.
Someone else pointed out that there are other ways to run Windows without Bootcamp.
But, as person number 1 pointed out, these days, many virtualisation setups use Bootcamp to power the Windows install, so the number of users who actually need to run Windows on a Mac is not that far from 2%.
There's a Visual Studio for the Mac??
It's wishful thinking basically. Whenever Apple does something like this, some people just take it personally. And when you read back on some of the positions they took earlier on, I think the problem is they feel Apple has made them look a bit foolish, especially those who claim to know Apple inside out.
https://forums.appleinsider.com/discussion/comment/3214174/
And looking at this:
Perhaps they think developers still write in machine code?