What happened to Intel? They seem to be in the same funk they were in back in 2005. They came out with the Core chips that helped break them out. Oh yeah, will Microsoft OS's run on the VM? I'd love to get my Servers over there as well. I fell asleep (long night - also getting sick...)during the presentation.
Intel is having a manufacturing problem - they have been stuck on 14nm for too long. That is not related to the inside of the silicon. Apple could have moved to AMD Ryzen as they are produced using the same 7nm by TSMC which also does the A-series, or Apple could have made a co-processor to deliver most of the features.
We didn't see any benchmarks comparing Apple Silicon to Intel and we didn't get any info on power consumption. It was just "this can run on Apple Silicon in 1080p" without showing frame rates.
Ryzen is x86, so that explains why Apple didn't go to the trouble of shifting from Intel earlier, given that their internal ARM roadmap gives Apple an advantage. At roughly the same 7nm node, Apple is almost certainly more power efficient than Rizen running x86, but benchmarking developer hardware today isn't going to be meaningful one way or the other. All that matters is that the developer hardware is fast enough to allow transition of apps to production Macs starting at the end of the year.
The new MacBooks must bring touch screens. The interface and iOS compatibility features are screaming for touch.
Mac Pro users lost again - if lucky they might be able to reuse their wheels
Linux compatible but no word on Windows?
If Messages is the most important Mac app and Safari is mainly for Pinterest... then this marks then end of the Mac for corporate use. Tim said PERSONAL computer. Don't expect docks or any focus on knowledge workers beyond group chat. iOS 14 uses AI to organize apps - but macOS doesn't use AI to organize documents or mix them with side panels showing related mails, chats,... that is super disappointing.
Don't get why Apple will launch more hardware using Intel. I don't even get how they will they sell their existing "built on a terrible platform with no future" products?
Until a few months ago I was thinking the same. There might be two outcomes in corporate use: either corporations would stick to their legacy systems centered on Windows and would make BootCamp a requirement for their Mac purchase,
OR,
the corporations might have just said "we buy Windows boxes a dozen a dime, we don't need Windows-compatible Macs, just give us pure native Macs". Obviously a Macbook Pro is too expensive to run Windows. Given the penetration of the iPhone and iPad into the corporate world, the trend may have chosen the second path. Obviously it is Apple who owns the data about that. Considering that Apple sells to the corporate world by ten thousands, it is unlikely that they didn't calculate the corporate sales loss when switching to Apple Silicon. They will continue to support the first path (BootCamp) with new Intel Macs inline, but in a couple of years those too may fade away depending on the penetration of Apple Silicon. Competition is good
Sad! grab an Intel Mac Pro while you can. ARM Macs will be inferior in performance.
They just demoed Maya running in emulation on a year-old chip designed for a thermally-constrained, fanless enclosure. MacBook Pros and Mac Pros will have chips designed for MacOS and Apple's compiler. Also imagine how powerful a chip they could create with active cooling. I expect they will destroy Intel in performance.
Unfortunately my MacBook goes obsolete too this year but I'm hoping to get a few more years use out of it.
Don't expect any improvements or future for the Intel based computers - beyond macOS 11.0. In October 2005 Apple introduced an iMac G5 using 10.4.2. It got killed after less than 3 months. The last OS for it was 10.5 - so just 1 OS update! Buying a Mac now is toxic!
WWDC was a very boring and a major disappointment because no hardware was announced as rumored! I fell asleep watching this 1 1/2 hour of the Craig Federighi stand up comedy audition! I do not trust anything this software engineer claims because anything Apple tries with software always does not work in the beginning until several updates later! Time to look at another computer brand because Apple is going to destroy their already limited line up!
The potential here is amazing. Apple will truly be the first company creating their own computers for the first time, designing everything under one roof, so they can innovate in brand new ways.
Every Apple event can mean major changes to the industry. They can start making computers with faster chips than anything else that's available. And, Intel will have to start innovating again, so better chip competition and faster chips there. But Apple will have the advantage, building their entire system on the newer technology of the two.
And like they said in the keynote, every Mac can have a great GPU now. Even the Mac Mini and Macbook Air can have a legit GPU in it, not just the Pros.
Plus, Mac will be able to leverage a huge base of info in the iOS ecosystem. The chips from the iPhone will subsidizing the entire operation, and iPhones are probably the most profitable 'computers' in the world. So, no longer will Mac be trying to keep up with off the shelf parts, but will be a first class member of the biggest show in town in it's little brother the iPhone.
For years, every time a new Mac is introduced, the peanut gallery whines that it's not based on the absolute latest or almost upcoming intel chip. This was probably because there was necessarily a lag time between Intel's development process and Apple's ability to take that and reliably integrate it into a machine that they're designing in concert with the next OS, etc. The whole point of Apple's business model is to build a seamless final product, not something that tries for bragging right on all fronts because jammed the newest off-the-shelf hardware together, untested, all into one box. So with the processor literally being the heart of the machine, buying it from Intel or somebody else creates a bit of a work-flow challenge. It makes sense that they'd want to bring that function back in-house.
I remain skeptical. Not because I don't think Apple can pull this off. They've done it 2x before so clearly they can. Consumers probably won't notice any difference other than better battery life, etc. Will large corporations buy any laptop that doesn't have an x86 compatible processor? Clearly they buy lots of iPhones and iPads so perhaps. Fortunately for Apple the Mac is a small portion of their overall revenue so the risk is relatively small. Especially compared to their last major transition when Mac was the majority of their revenue. This is a multi-year adventure so it will be interesting to follow Apple's progress.
I see no reason a large corporation would care about x86 compatibility for a Mac.
Unlike what the article says, Apple didn’t say what the machines they were using for their demos had in the way of RAM or storage. They did say how much would be in the developer machines. And I would also like to point out that the iPad Pro has 6GB RAM, not 16GB, so the developer machines are not outfitted the same. That’s not surprising, because the Mac needs more resources than iOS devices.
When Craig pulled up the About This Mac screen it said 16 GB and he said it was the machine they were doing the demos on. Not sure what else you need.
Sad! grab an Intel Mac Pro while you can. ARM Macs will be inferior in performance.
*sigh*
We just saw the Intel version of Maya running on an A12Z-based Mac. That Mac is just a Developer Kit - it is not in anyway indicative of the the SoCs Apple is designing specifically for the Mac.
Maya’s own minimum recommendations for running their software is a 3.6GHz 8-core i7 and 16GB of RAM.
This looks great, and that Tomb Raider Demi in Rosetta? Wow. Can’t wait for these things to take off. Hopefully they do a beefy update to the mini soon after launch.
I was a bit worried when they mentioned the same iPad chip until I saw that demo. Running Shadow of the Tomb Raider at 1080p that smoothly while being emulated is really impressive. Anandtech did a test in 2018 of the A12X chip:
The Nvidia 1060 notebook GPU is around 4TFLOPs and the A12X is approx 60% of that. A12Z scores around 220,000 (67%). This would be just a bit below the 5300M GPU in the base MBP around the power profile of an iPad.
CPU-wise A12Z performs like the 6-core cylinder Mac Pro and 6-core MBP.
Even if the benchmarks are a bit off, the power profile of a Mac mini is 6x an iPad, a 16" Macbook Pro is 8x.
Although it's the same chip design, they can run the clock speeds higher and keep it actively cooled.
This may lose Bootcamp but they have custom hardware and they can build better virtualization tools that allow better GPU sharing. In the end it comes down to performance and compatibility. I don't expect this will cover 100% of the use cases on Intel but it will mean more efficient hardware and hopefully cheaper hardware so worst case for the cases not covered, a cheap Windows machine can be bought for that.
Xcode is now running on ARM and will finally get full hardware acceleration when testing mobile apps:
Things will become clearer once the hardware has been tested but it looks good so far.
It looks like they will start at the low-end. These are machines that are hampered by poor Intel IGPs and make up the large majority of Apple's Mac userbase (70%+). Over the next 2 years when Apple starts expanding the chip line and moves to 3nm, it will make sense to move the higher end models too.
If the road this ends up going down means the same apps on a Mac are the same apps on an iPad, then surely Apples PC market share is going to shrink even further, everyone will just buy a much cheaper iPad.
I’m not questioning Apples silicone prowess here, I’m questioning if developers will follow them down the path.
That isn’t how apps are developed. Apple are not changing the silicon so they can write the same apps for both, as was explained in the Keynote. They are doing it for a lot of other reasons. Thinner laptops being one I’d imagine!
Apple already have Safari, Keynote, Photos, Mail, Maps, etc, etc, etc running on both platforms. They are not the same apps and putting the same processors in each won’t change that at all.
It will take effort, money and resources for a developer to make their programme that runs on X86 PCs also run on Mac OS Apple Silicone machines, and when Apple has a single digit market share, how many are going to do that? Even more so if they have an iOS app I think most will just let that be the programme for the new Macs instead, better value for the company and those share holders, Mac users get a lesser experience though.
A lot of apps will simply need to be recompiled. Some will need more. Almost all apps are written for an OS, not a chip. That’s why we have compilers - they take the C, C++, Swift code or whatever and turn it into code that targets a particular processor. Unless you do bit manipulation and some other clever stuff I’ve had no experience off, the processor in the computer is irrelevant when writing an app.
the Swift code I write is the same right now whether it’s designed to run on a Mac or iPadOS- the Swift language that is, not the libraries which ARE different.
for example. Download Xcode for your Mac and if you have an iPad, Swift playgrounds for iPadOS. Write swift code to create variables, functions, structs and more in a Swift playground on the Mac. Run it and see the result. Select your Swift code’s text and copy it into the playground on your iPad. Run it. It will do exactly the same thing. Two different OSs, one Intel, one A series. You didn’t write for a processor, you wrote Swift. That’s how it goes.
i have a Mac App on the Mac App Store. When Big Sur is released I will just launch it in Xcode, Build, archive, upload to the App Store and bingo, I will have an app that runs on an A series chip. That, for many developers, is all that is needed.
I thought it was interesting that the transition strategy was an exact mirror of 2005, with virtualization added. This will be far easier, though, with Carbon long dead and all apps 64 bit. No 2-year wait for native Photoshop.
WWDC was a very boring and a major disappointment because no hardware was announced as rumored! I fell asleep watching this 1 1/2 hour of the Craig Federighi stand up comedy audition! I do not trust anything this software engineer claims because anything Apple tries with software always does not work in the beginning until several updates later! Time to look at another computer brand because Apple is going to destroy their already limited line up!
If Rosetta is generating a translation of x86 apps on install, does that mean the storage for those apps is going to balloon (until they get recompiled, obvs)?
Yes, this was the worst kept secret in the industry, just like the PowerPC -> Intel transition. But there are some interesting long term ramifications.
First, the fact that more Intel Macs are coming is testament that Apple thinks they can transition the Mac to AppleSilicon and the vast number of users not even notice. It's just a compile and some QA. Apple usually supports macOS for at least 6 years for most hardware and the Mac Pros usually get 8. So there will be Intel versions of macOS long after the last Intel Mac is sold new.
Pundits have talked about the fact that the current iPad processor is quite competitive with Intel hardware and that's the reference hardware developers will be working from. Do you think that is going to be the speed/performance of the first shipping Macs? I think not. The first machines will have to a make a statement that Apple can do this better than the platform they are leaving behind. If it's indeed a "Pro" machine, then that will be a really big deal.
Apple has already slain the power consumption dragon by a decade of iPhone chips. What will be interesting is Mac chips. On the slide they showed, they indicated there will be high performance AND high efficiency cores. How many of each for a typical Mac is anyone's guess. But I'd wager that they way Apple configures a chip might be different than anything currently available.
The presentation also put a stake in the heart of AMD in that Apple may not need them for advanced graphics either, but that remains to be seen. We've already seen the Mac Pros accelerator card which is probably a bunch of custom design GPUs. That kind of architecture is probably just the beginning. Someone earlier questioned if Apple will have advanced ray-tracing like Nvidia and AMD are doing now. Hard to say, but i wouldn't bet against Apple.
I/O will be interesting to see play out. Any new AS Mac will need things like Thunderbolt, HDMI (Mac Mini), and USB. All of that will be in the reference developer machine. Will AS Macs debut with Thunderbolt 4 which will unify Thunderbolt + USB? The timetable of both toward the end of this year will be interesting. The slide during the presentation spoke of a common memory architecture, which i read as common from CPU and GPU. This is standard practice in phones and tablets and we saw it in mainline computers when Intel gave us GPUs in Core i3/i5/i7 processors. But Intel took awhile to get GPU performance up to snuff. I don't see Apple making that mistake so the common memory will be a key performance variable. But Pro GPUs use more exotic memory in order to keep pipelines stuffed. How will Apple solve this?
Some of said that Mac Pro users will be pissed? Why? Those machines were built for a job....to make money. Every one in production right now is chewing up watts earning their keep, and will for years. The real test will be a "server" version of the A-series processors that a Mac Pro or iMac Pro can use that will kick the crap out of Xeon or Ryzen. Back in 2006, we got the first Intel Mac Pro whose hardware was based on the previous G5 PowerMac. I think the hardware basics will remain the same, but the CPU guts will look very different. And graphics? Too soon to tell.
From a marketing standpoint, will the shipping Mac chips be A-series chips or will they have their own designation? Maybe the T3? The T2 is already a specialized ARM-based chip....why not expand it to do everything and eliminate a chip from the board?
How do you conclude that MacPro users would be lost? The MacPro has this magical thing called a PCIe slot. Have an Intel MacPro? Get a PCIe card with Apple Silicon. Have a MacPro with Apple Silicon? Get a PCIe card with Intel x86. This is a non-issue.
There is no PCIe card with Apple Silicon and there won't be. Think security on a card + power + system integration + the fact that there are very few sold units. On the Mac Pro storage is connected to T2. And it won't work with the existing GPU in the Mac Pro. Last time Apple dropped computers launched 2-3 months earlier, and those computers got one OS update (10.4.2 to 10.5). The Mac Pro is toast - again. In the old days Apple offered motherboard replacements but that stopped like 20 years ago.
The Mac Pro is PCIe 3.0 - which simply isn't fast enough. Even budget computers from AMD are now running PCIe 4. The Mac Pro 2019 was built using tech that was obsolete on launch. You can get a B550 motherboard with a PCIe 4.0 SSD for 50-100% better performance on storage.
Maybe I missed it, but I haven't heard them say ARM once. I wonder if that means anything.
ARM means: Acorn RISC Machine. Acorn was the grand father of the ARM RISC CPU design. They had their own operating system as well as computers. A lot has changed since then and a lot of Apple’s additions and changes aren’t based on that classic ARM architecture anymore. Apple will prefer to use their own brand name compared to regering to a legacy name like ARM.
Best WWDC keynote in years. Straight to the point, no interacting with a life audience, pre-recorded is they doctor get it right without hiccups and keep it tightly scripted. They should do it this way every year, with apologies to the live audience who would love to be present.
I thought it was overproduced, sterile and lifeless. And the typical American superlative Ken doll acting. Cringe. I did like however what they announced, an amazing update.
How do you conclude that MacPro users would be lost? The MacPro has this magical thing called a PCIe slot. Have an Intel MacPro? Get a PCIe card with Apple Silicon. Have a MacPro with Apple Silicon? Get a PCIe card with Intel x86. This is a non-issue.
There is no PCIe card with Apple Silicon and there won't be. Think security on a card + power + system integration + the fact that there are very few sold units. On the Mac Pro storage is connected to T2. And it won't work with the existing GPU in the Mac Pro. Last time Apple dropped computers launched 2-3 months earlier, and those computers got one OS update (10.4.2 to 10.5). The Mac Pro is toast - again. In the old days Apple offered motherboard replacements but that stopped like 20 years ago.
The Mac Pro is PCIe 3.0 - which simply isn't fast enough. Even budget computers from AMD are now running PCIe 4. The Mac Pro 2019 was built using tech that was obsolete on launch. You can get a B550 motherboard with a PCIe 4.0 SSD for 50-100% better performance on storage.
PCIe-4 support depends on Intel, not on Apple. Show us any Xeon that supports PCIe-4 yet. Your point is irrelevant.
The new, Mac specific Apple SOCs are going to be incredible.
They’ll retain the efficiency of the iPad while gaining headroom for performance and capability.
Much thinner computers on the way, which is pretty neat I guess from an aesthetics perspective, but it sounds like there is a good possibility thst new Macs will dramatically outpace all other PCs, even with the best Intel chips.
And Apples economy of scale and sharied architectures will allow for the price to decrease as well if Apple chooses to pass the savings on to the consumer.
And then there is always the possibility of multiple processors, each with multiple cores, managed by a hardware controller (which Apple could also mass produce as a separate SOC if desired.
The performance possibilities are exciting.
Can’t wait for Augusta/September for new hardware.
Rosetta 2... “recompiling” the app DURING INSTALLZ! a BRILLIANT! rather than have Mac OS translating in the fly, just front Load the delay during install and let it fly during use. This is how it should be done.
Apple is starting to look like Apple again. Loving it.
Mac OS redesign.
Really love the unification in aesthetics and looks very Ive inspired, but is adding back the subtle little “cute” touches that were long-standing examples of Apples heartfelt attention to detail.
Mac and I devices running the same apps is so right.
No more slienationnjust because you need to get some serious mouse and keyboard work done.
I imagine touchscreen will be a big across the board Mac feature as well.
The Mac is finally getting parity attention with the mobile devices. It’s about time.
Keeping the fundamentals of what makes a computer a computer, but improvements worthy of the eta and beyond.
Comments
Ryzen is x86, so that explains why Apple didn't go to the trouble of shifting from Intel earlier, given that their internal ARM roadmap gives Apple an advantage. At roughly the same 7nm node, Apple is almost certainly more power efficient than Rizen running x86, but benchmarking developer hardware today isn't going to be meaningful one way or the other. All that matters is that the developer hardware is fast enough to allow transition of apps to production Macs starting at the end of the year.
OR,
the corporations might have just said "we buy Windows boxes a dozen a dime, we don't need Windows-compatible Macs, just give us pure native Macs". Obviously a Macbook Pro is too expensive to run Windows. Given the penetration of the iPhone and iPad into the corporate world, the trend may have chosen the second path. Obviously it is Apple who owns the data about that. Considering that Apple sells to the corporate world by ten thousands, it is unlikely that they didn't calculate the corporate sales loss when switching to Apple Silicon. They will continue to support the first path (BootCamp) with new Intel Macs inline, but in a couple of years those too may fade away depending on the penetration of Apple Silicon. Competition is good
In October 2005 Apple introduced an iMac G5 using 10.4.2. It got killed after less than 3 months. The last OS for it was 10.5 - so just 1 OS update!
Buying a Mac now is toxic!
*sigh*
We just saw the Intel version of Maya running on an A12Z-based Mac. That Mac is just a Developer Kit - it is not in anyway indicative of the the SoCs Apple is designing specifically for the Mac.
Maya’s own minimum recommendations for running their software is a 3.6GHz 8-core i7 and 16GB of RAM.
https://www.anandtech.com/show/13661/the-2018-apple-ipad-pro-11-inch-review/6
https://www.notebookcheck.net/A12Z-Bionic-GPU-vs-A12-Bionic-GPU_10326_8896.247598.0.html
The Nvidia 1060 notebook GPU is around 4TFLOPs and the A12X is approx 60% of that. A12Z scores around 220,000 (67%). This would be just a bit below the 5300M GPU in the base MBP around the power profile of an iPad.
CPU-wise A12Z performs like the 6-core cylinder Mac Pro and 6-core MBP.
https://browser.geekbench.com/ios-benchmarks
https://browser.geekbench.com/mac-benchmarks
Even if the benchmarks are a bit off, the power profile of a Mac mini is 6x an iPad, a 16" Macbook Pro is 8x.
Although it's the same chip design, they can run the clock speeds higher and keep it actively cooled.
This may lose Bootcamp but they have custom hardware and they can build better virtualization tools that allow better GPU sharing. In the end it comes down to performance and compatibility. I don't expect this will cover 100% of the use cases on Intel but it will mean more efficient hardware and hopefully cheaper hardware so worst case for the cases not covered, a cheap Windows machine can be bought for that.
Xcode is now running on ARM and will finally get full hardware acceleration when testing mobile apps:
https://developer.apple.com/documentation/metal/developing_metal_apps_that_run_in_simulator
Things will become clearer once the hardware has been tested but it looks good so far.
It looks like they will start at the low-end. These are machines that are hampered by poor Intel IGPs and make up the large majority of Apple's Mac userbase (70%+). Over the next 2 years when Apple starts expanding the chip line and moves to 3nm, it will make sense to move the higher end models too.
the Swift code I write is the same right now whether it’s designed to run on a Mac or iPadOS- the Swift language that is, not the libraries which ARE different.
for example. Download Xcode for your Mac and if you have an iPad, Swift playgrounds for iPadOS. Write swift code to create variables, functions, structs and more in a Swift playground on the Mac. Run it and see the result. Select your Swift code’s text and copy it into the playground on your iPad. Run it. It will do exactly the same thing. Two different OSs, one Intel, one A series. You didn’t write for a processor, you wrote Swift. That’s how it goes.
i have a Mac App on the Mac App Store. When Big Sur is released I will just launch it in Xcode, Build, archive, upload to the App Store and bingo, I will have an app that runs on an A series chip. That, for many developers, is all that is needed.
Buh Bye!
Yes, this was the worst kept secret in the industry, just like the PowerPC -> Intel transition. But there are some interesting long term ramifications.
First, the fact that more Intel Macs are coming is testament that Apple thinks they can transition the Mac to AppleSilicon and the vast number of users not even notice. It's just a compile and some QA. Apple usually supports macOS for at least 6 years for most hardware and the Mac Pros usually get 8. So there will be Intel versions of macOS long after the last Intel Mac is sold new.
Pundits have talked about the fact that the current iPad processor is quite competitive with Intel hardware and that's the reference hardware developers will be working from. Do you think that is going to be the speed/performance of the first shipping Macs? I think not. The first machines will have to a make a statement that Apple can do this better than the platform they are leaving behind. If it's indeed a "Pro" machine, then that will be a really big deal.
Apple has already slain the power consumption dragon by a decade of iPhone chips. What will be interesting is Mac chips. On the slide they showed, they indicated there will be high performance AND high efficiency cores. How many of each for a typical Mac is anyone's guess. But I'd wager that they way Apple configures a chip might be different than anything currently available.
The presentation also put a stake in the heart of AMD in that Apple may not need them for advanced graphics either, but that remains to be seen. We've already seen the Mac Pros accelerator card which is probably a bunch of custom design GPUs. That kind of architecture is probably just the beginning. Someone earlier questioned if Apple will have advanced ray-tracing like Nvidia and AMD are doing now. Hard to say, but i wouldn't bet against Apple.
I/O will be interesting to see play out. Any new AS Mac will need things like Thunderbolt, HDMI (Mac Mini), and USB. All of that will be in the reference developer machine. Will AS Macs debut with Thunderbolt 4 which will unify Thunderbolt + USB? The timetable of both toward the end of this year will be interesting. The slide during the presentation spoke of a common memory architecture, which i read as common from CPU and GPU. This is standard practice in phones and tablets and we saw it in mainline computers when Intel gave us GPUs in Core i3/i5/i7 processors. But Intel took awhile to get GPU performance up to snuff. I don't see Apple making that mistake so the common memory will be a key performance variable. But Pro GPUs use more exotic memory in order to keep pipelines stuffed. How will Apple solve this?
Some of said that Mac Pro users will be pissed? Why? Those machines were built for a job....to make money. Every one in production right now is chewing up watts earning their keep, and will for years. The real test will be a "server" version of the A-series processors that a Mac Pro or iMac Pro can use that will kick the crap out of Xeon or Ryzen. Back in 2006, we got the first Intel Mac Pro whose hardware was based on the previous G5 PowerMac. I think the hardware basics will remain the same, but the CPU guts will look very different. And graphics? Too soon to tell.
From a marketing standpoint, will the shipping Mac chips be A-series chips or will they have their own designation? Maybe the T3? The T2 is already a specialized ARM-based chip....why not expand it to do everything and eliminate a chip from the board?
A lot has changed since then and a lot of Apple’s additions and changes aren’t based on that classic ARM architecture anymore.
Apple will prefer to use their own brand name compared to regering to a legacy name like ARM.
I did like however what they announced, an amazing update.
The performance possibilities are exciting.