I think you are missing the point here....the GPU is built in, that is why apple is building its own as it does both functions very well....
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Transitioning Macs to Custom Chips With 'Incredible' Performance Starting Later This Year
- Thread starter MacRumors
- Start date
- Sort by reaction score
I think you are missing the point here....the GPU is built in, that is why apple is building its own as it does both functions very well....
RISC has an inherent advantage when dealing with memory because you can coalesce reads and writes (since only load/store instructions access memory, not every ALU instruction, etc.) and because you have more registers (so you need to read or write to RAM less frequently).
.... I am looking at 2 24" screens right now. Do you realise how big 24" is? Nobody is going to carry that ****er around - they couldn't even sell enough 17" laptops to make them a worthwhile product line.
Also, let's assume for a minute Tim Cook becomes a size queen and does announce a 24" iPad. How would it not be a touch screen? It's literally the primary source of input - and for most people, the only source of input.
I know you’re trying to make assumptions based on my terse message, but when I said “full stack” that meant I use Microsoft technologies, including .NET framework which is not available on Visual Studio for Mac. It will reach legacy status eventually but I have to support it in the meantime all the same.
When I say “primary machine” that means the tower used for development. I can still use a Mac laptop for testing browsers, although I frequently have to work remotely so an ARM‘ed Mac without access to Windows won’t work in that case either.
The short-term solution is to buy the final generation of Intel Macs (desktop and laptop) to guarantee Windows compatibility for the foreseeable future. (My current machines are 8-10 years old so purchases are due.) However instead of a knee-jerk reaction purchase, I’m giving Microsoft a bit of time to see if they will commit to a proper 64-bit version of Windows 10 for ARM, as well as Apple for giving a solution to the loss of x86 Windows. It‘s likely a small target market, but I’m part of that market so that’s how it has to be.
I would be more than happy to buy an ARM’ed Mac if there were solutions for these concerns, but I need to purchase what works for my clientele.
Last edited:
As much as it pains me to have to go through this transition, I am firmly convinced that in the long term Apple will be much stronger as a consequence of it.
Apple strength has always come from being in control of every aspect of their hardware and software ecosystem.
They are now in a position where they can take the last step to obtain full control.
Apple strength has always come from being in control of every aspect of their hardware and software ecosystem.
They are now in a position where they can take the last step to obtain full control.
Huh, I can't say I ever considered someone elbow deep in .NET/etc would use a Mac - it sounds like you're not just using the VM as a deployment target for a local dev server, but as a full GUI to run the full-fat version of Visual Studio? Or are you doing this via Bootcamp and boot back to macOS for non-.NET work?
In the ~13 years I've had Intel Macs, I don't think I've ever used a windows (or Linux for that matter) desktop (i.e. with a GUI) VM and thought "hey yeah I could use this all the time". It was always like cleaning the gutters standing on a ladder. Yes, it needs to be done, so you have to use the ladder. But you're not gonna just choose to stand on a ladder all the time if you have an alternative. (Weird analogy I know - I couldn't think of anything else that is sometimes required to get the job done but that most people would get uncomfortable doing all the time).
Rather like Windows-on-Windows. It takes care for ABI differences of different architectures when it comes to system DLLs.
Thats what i said, didn't I?
Inform yourself! Windows RT had native Office. Now for 64bit Windows on ARM, they changed the concept - due to inherent problems with existing plugins. Instead they developed CHPE libraries, which are essentially ARM64 DLLs with a small x86 ABI wrapper. Office is using this concept, such that only a very small application layer needs to be emulated. Most of the work is done in native CHPE libraries.
I did not claim anything else. Besides WSL2 is quite a bit more than a VM running a Linux Kernel.
Read my explanation from above. It is save to assume that developing the CHPE concept did cost much more effort compared to compile Office just for ARM, as they did for Windows RT.
ARM ATX boards already exist. Clover can be adapted and ported. Hackintosh will still live.
Last edited by a moderator:
The last transition was much harder for several reasons:
- There was a difference between the endian-ness of the machines that meant that were some issues with low level code written for performance and for drivers.
- Many developers were not using Apple tools/compilers for their software.
Think about how long it took for Microsoft and Adobe to have truly native Intel products last time. This time, I bet they will both ship at launch.
In addition, the last transition happened at a point where the Mac was a much smaller market. The Mac is much healthier now, so I expect it to be much easier.
This time we expect real benefits, lots of them. Including:
- Custom silicon with functions optimized for the OS (hardware enhanced video compression/decompression, Secure Enclave, low power cores for updating while mostly sleeping, etc.).
- A product timeline completely controlled by Apple, no longer being dependent on Intel or AMD (look at the steady march of Apple‘s SoC’s in the phone space, relentless, year after year of meaningful performance gains).
- Better per watt performance (the primary reason for the last transition), making possible products that run cooler and quieter, with better performance and battery life.
- Compatibility with the large variety of iPadOS apps (actually, a much larger market than the Mac).
I am curious about three things:
- When did you purchase your last Mac?
- Were you about to upgrade?
- Do you run Windows on your Mac as a primary use case (either as a developer, gamer, or for specific Windows only apps you use on a daily basis)?
My son will be attending school remotely in the fall, so I’m considering an ARM-based Mac for him for this fall (or winter). He didn’t have a computer until the COVID-19 pandemic hit, and when the schools closed, I gave him my old 2012 Mac Mini for his remote school lessons. Now he started streaming Xbox One NHL game from the downstairs Xbox One to the Mac Mini in his room by booting into Windows 10 on his Mac.
If I wanted to get him an ARM-based Mac to replace his current 2012 Mac Mini, would he be able to game on his computer in his room by streaming his NHL games from the Xbox One sitting in the living room? Somehow I doubt it. Mind you, this is not an enterprise requirement, but the ARM-based platform already presents a limitation in this simple use case scenario.
His school doesn’t require a Mac. Any computer works for the live online lessons, even ChromeOS works just fine. Guess what computer will serve all of his computing and gaming needs going forward? Not an ARM-based Mac. He will probably end up with a Dell XPS.
My point is this: transitioning to ARM is a good thing. Dropping Windows for X86 is a disaster - both for consumers and for businesses. But then again, Apple really couldn’t care less. Macs are but a small drop in their corporate profit bucket.
If I wanted to get him an ARM-based Mac to replace his current 2012 Mac Mini, would he be able to game on his computer in his room by streaming his NHL games from the Xbox One sitting in the living room? Somehow I doubt it. Mind you, this is not an enterprise requirement, but the ARM-based platform already presents a limitation in this simple use case scenario.
His school doesn’t require a Mac. Any computer works for the live online lessons, even ChromeOS works just fine. Guess what computer will serve all of his computing and gaming needs going forward? Not an ARM-based Mac. He will probably end up with a Dell XPS.
My point is this: transitioning to ARM is a good thing. Dropping Windows for X86 is a disaster - both for consumers and for businesses. But then again, Apple really couldn’t care less. Macs are but a small drop in their corporate profit bucket.
Last edited:
You would be correct, sir. I usually use VMWare but sometimes performance is dismal enough that logging in natively with Bootcamp works much better. I also play the occasional PC game on Steam which also needs native boot support. All web dev work and gaming is on the PC side, all other work (music/video/photo/finance/administrative) is on the Mac side. I'm your typical edge case use.
I've been on Macs since the early 1990s, so could never give them up completely. Prior to the Intel switch I had a PowerMac 6500 and Pentium III side by side and it was annoying. The Mac Pro was the one stop shop answer I had been waiting for. I will stretch that out as long as possible if there isn't an immediate Windows solution for this ARM transition.
technical superiority is cool and will benefit Apple, particularly in the affluent consumer space - the technical aspects are exciting. I will likely buy an arm MacBook as a superiPad.
- fortunately for Intel there remains a massive market for x86 processors to run professional software. AMD is better but it’s overkill for much of this professional software (I just bought an 8 core zen 2 and the performance difference for work is barely noticeable over my 5 years old 6 core 12 thread intel). With a fast ssd and a decent amount of ram hardware is not the barrier from making it a great pc.).
up until zen 2 amd had compatibility issues with ram. Intel “just works”.
arm mac will be like my tesla 3 - superior to an ice Vehicle (x86) in every way. But there are plenty of use cases for an ice vehicle And there will be for some time.
Windows 10 for ARM was released in 2017.
You are probably thinking of Windows RT which has nothing to do with anything. RT was Windows failed equivalent to iOS.
Anyway, I am not here to defend Microsoft. The original point was to say that there is a native ARM version of Windows 10 and we have established that the answer is a resounding yes.
If you want to run Windows with great performance, you will be better off buying a PC rather than worrying as to whether it can be run on a Mac. This is true whether your Mac is powered by Intel or whether in the future will be powered by Apple's ARM SoC.
It will be an interesting transition, and it was announced just 15 years after Apple announced the PowerPC to Intel transition. Impeccable timing!
Once I get me an I.T.-related job (hopefully this summer!) I am going to want to buy a new Mac laptop to replace my late 2009 polycarbonate MacBook, but I might just go for a 13" Touch Bar MacBook Pro that still has an Intel quad-core processor. I'm going to wait for a year or two to get an Apple Silicon processor Mac, probably once they've come out with a Mac Mini using such a processor and I can replace my 2012 quad-core i7 Mini. (The MacBook Pro will come first, to replace my 10 1/2-year-old MacBook and so I can at least have one computer that can run Mac OS 11 Big Siri, while I continue to use my 2012 Mac Mini until I'm ready to buy an Apple processor Mini.
I also suspect this will also mark the end of the line for Boot Camp, but I don't really use that anymore, not since I got one of my dad's Dell Optiflex 9010 towers running Windows 10 from his old job so I can have two different desktop types set up in my room, since I mentioned I do enjoy being a cross-platform user, especially since I'm going to work in I.T. But I DO know the demonstration showed Parallels Desktop running on that Apple Silicon Mac prototype, and even though not mentioned, I can safely assume we can still run Windows virtual systems on Macs with the new chip via Parallels or VMWare Fusion or VirtualBox. (This would be good, as on my Mac Mini I do sometimes use a Windows XP virtual machine in VMWare Fusion for throwback purposes, especially if I am making a "Let's Play" commentary video on one of the computer games of my childhood!)
Once I get me an I.T.-related job (hopefully this summer!) I am going to want to buy a new Mac laptop to replace my late 2009 polycarbonate MacBook, but I might just go for a 13" Touch Bar MacBook Pro that still has an Intel quad-core processor. I'm going to wait for a year or two to get an Apple Silicon processor Mac, probably once they've come out with a Mac Mini using such a processor and I can replace my 2012 quad-core i7 Mini. (The MacBook Pro will come first, to replace my 10 1/2-year-old MacBook and so I can at least have one computer that can run Mac OS 11 Big Siri, while I continue to use my 2012 Mac Mini until I'm ready to buy an Apple processor Mini.
I also suspect this will also mark the end of the line for Boot Camp, but I don't really use that anymore, not since I got one of my dad's Dell Optiflex 9010 towers running Windows 10 from his old job so I can have two different desktop types set up in my room, since I mentioned I do enjoy being a cross-platform user, especially since I'm going to work in I.T. But I DO know the demonstration showed Parallels Desktop running on that Apple Silicon Mac prototype, and even though not mentioned, I can safely assume we can still run Windows virtual systems on Macs with the new chip via Parallels or VMWare Fusion or VirtualBox. (This would be good, as on my Mac Mini I do sometimes use a Windows XP virtual machine in VMWare Fusion for throwback purposes, especially if I am making a "Let's Play" commentary video on one of the computer games of my childhood!)
What was demonstrated was a Debian Linux Arm64 VM. Apple has clearly stated that they will not support emulation for Intel VM, so it looks like the only thing possible might be Windows for ARM, but so far no one knows.
I think the Mac gaming community is about to get turned on it’s head to be honest. Apple is clearly building some great graphics going forward.
Obviously that will do nothing for or existing games (especially on Windows).
For once the timing is going to work out for me. I have a 2014 iMac 5K that I was planning on replacing within the next 12 months. I will definitely wait to get the new ARM based iMac.
I can't say for sure what interface is used on the iPad motherboard for the NAND storage. What I'm pretty sure is, the peripheral access is handled by the I/O included on the system on a package and then the controller talks to the CPU. PCI's lanes requires direct access to the CPU.
Someone with a better understanding can address these questions.
A computer is not an investment, it is an expense. One can be pretty sure that these machines will be well supported for at least 3 years. If one needs a machine to last for longer than that, it might make sense to wait until the fall. If one replaces a machine every two to three years, then it really depends on how far over three years one is right now. In general, waiting until one’s current machine does not meet one’s needs usually makes sense, as the tech keeps moving.
It can run an instruction every clock cycle and so can CISC. With today's microprocessor technology the benefits of RISC have all been but render moot.
[automerge]1592931559[/automerge]
1) What is "is Load/store architecture"? What does that mean?
2) Already addressed. An orthoganal instruction set may have had benefit in the early days of processors. Today the instruction decode logic is a small part of a processor design as to have rendered this potential benefit moot.
3) The number of registers is not a function of RISC or CISC. It's a function of how many registers the processor is designed with.
4) Weakly order memory model? What does this mean?
5) How does this benefit RISC? As already mentioned processor technology has all but rendered this moot.
No they haven’t. You can’t explain why somehow having a ton of extra pipe stages, a much bigger instruction decoder, a much higher branch prediction miss penalty, far fewer registers, etc. etc. are all “moot.”
[automerge]1592931636[/automerge]
Yep.
Hell, did they even stop making itaniums yet? Just recently maybe?
i doubt this as i mentioned above. apple's chips are SoCs with their own memory controllers, pcie controllers, graphics chips, etc. if all the code in OSX is centered around initializing and abstracting these chips, it will be a ton of work, and quite possibly impossible, to port OSX to a machine with a completely different set of hardware, without the source code to OSX.
on iOS the bootloader has to decrypt the image and apple's key is (still) secret. if ARM OSX is the same, you can forget about booting ARM OSX on anything else but apple hardware. clover and opencore will not know how to decrypt the binary, so it has nothing to hand off to.
Last edited:
Yes, they have. You can't (or at least haven't attempted to) explain why RISC is inherently better. You're very good at throwing around buzzwords but lack any explanation as to why these are problems.
Let's use "extra pipe stages" as an example. CISC doesn't require any specific number of stages, certainly no more than a RISC design. The number of stages are what the designers felt were appropriate for the specific design. The Netburst processors had a long pipeline. The design was to have a longer the pipeline so the processor could have a higher clock speed. It worked for a while until Intel hit power limits. Later designed used fewer stages and correspondingly slower clock speeds.
I also chose "extra pipe stages" because it also affects branch prediction misses. Longer pipelines have to discard more inflight instructions if a branch is taken. Therefore it can take more clock cycles to refill the pipeline from the new instruction sequence.
Given the transistor counts in a modern processor how does having a "much bigger" (whatever that means) instruction decoder hinder performance. The instruction decoder is but a small part of a modern processor. It's certainly not a hindrance to performance, at least not that anyone has demonstrated. Merely being "much bigger" doesn't make it slower. Even if it were a modern processor can load and decode multiple instruct at a time rending the instruction decode "advantage" moot.
Nope. Over time Swift UI will be the One True Kit (TM). That will make this all work well.