There are plenty of developers out there who develop on a Mac but don’t target the Mac platform per se.
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.
Bloomberg: Apple's First Arm Mac to Launch by 2021 With 12-Core Processor
- Thread starter MacRumors
- Start date
- Sort by reaction score
There are plenty of developers out there who develop on a Mac but don’t target the Mac platform per se.
I know that, but that's the risk they take. Apple does what benefits them - not a minority with edge case usage profile. That's how it's always been.
Sure, I can buy a Porsche 911 Convertible, remove the soft top, weld on some steel roof I made and take that thing to the race track. Or, I could just buy the 911 Coupé …
Adequate tool for the job.
Cinema 3d, 3dx Max, Rhino, ZBrush, Unreal
Then there are the render engines themselves.
Of course, Blender will show up (if someone wants to scratch that itch.)
How long until a feature compatible Adobe CS suite show up?
The macbook air lacks ports, so don't expect anything that hangs off the computer itself to show up anytime soon.
Until now, the Mac was an adequate tool for the job. Moving to ARM isn’t something that has been on everyone’s minds until these rumours started and Apple acquired a processor design shop.
You named some software, but I still don't see why these cannot translate to ARM.
You guys need to stop confusing "Hasn't been done yet" with "Cannot be done".
I‘m not saying it cannot be done, I’m asking what the timeframe is, given you think it’s easy as abc. The theory is easy, the practicality is that compiling for a new architecture comes with many things.
That is completely true. We don’t know when they started on that. But based off the article, it sounds like they have been focused on the updated A13 chip that will go into a MacBook, and then they plan on working on a new chip based off the A14 chip next year. Which again, sounds like they aren’t really planning on replacing the Mac Pro yet and haven’t really started on the other chips yet.
And as far as a time line, yeah I agree there is no way Apple will replace the Mac Pro in 2021 or even 2022 with ARM since they just made a new Mac Pro. So that will be the very last device to updated.
I expect - based off this rumor, of course - the first device we will see is a MacBook in June 2021. Then the Mac Mini and/or iMac in Fall 2021. Maybe both at the same time or maybe one first and then the next. If they stagger that release, we will see the next one in Winter 2022 (Dec 2021 - Feb 2022).
Then in June 2022 we’ll see the iMac Pro get an ARM update (and I’m assuming of course that the iMac Pro doesn’t get an Intel update during this time). And they will announce the final Mac Pro with its new processor at WWDC 2022 (which will be a beefed up version of the iMac Pro CPU). It will introduce the first time an ARM Mac will have interchangeable GPU’s just like the current Mac Pro. And that should come out at the end of Dec 2022.
That is three years since the Mac Pro was released, and both Mac Pro’s will still be supported for at least 3 years after this with software updates. Giving the Mac Pro Intel 6 years of updates which is about right.
Obviously all of this is an educated guess. It could be longer and the dates could be totally off. But I don’t expect it to be earlier then that and I dont expect this to happen over night.
Sure, but in the case of Docker, VMware, Parallels, and Boot Camp to a productive Windows environment there are plenty of tools which are in the "cannot be done" category.
If you insist on doing your own thing while coding instead of using the tools the OS vendor supplies you with, you might have a bit more work to do.
That is true. At least if you talk x86 Windows environment. But again, that's nothing Apple has to take into consideration. They did not switch to Intel to make that possible for their customers - it was a side effect. And even though I do get that for some people this is important, I believe it has very low to zero priority for Apple.
Yes. IDEs, such as Xcode, do a very seamless job of compiling, usually for a wide range of instruction sets. The few developers that would have a big hassle are those coding for drivers. But again, that's a very small percentage of programmers. Most developers rely on APIs. As also mentioned above, one of the biggest hurdles of software development is with libraries and frameworks When moving or adding platforms to your apps portfolio, you may need to utilize different libraries, frameworks because many are OS or platform-specific. Because Web browsers and the technologies they rely on/implement are mightily cross-platform, Web apps are one of the most resilient and flexible to develop, which is why they're so popular.
My quad-core i7, 32GB RAM, fast SSD laptop that's running Windows 10 natively can't even run a virtual copy of Windows 10 -- clean, simple install -- buttery smooth.
I confidently say Real Racing 2 HD on an iPad 2 or iPad 3rd generation rivaled consoles of its time.
I bet the RPi 4 config could trounce a Core-M MacBook :-D
Not everyone wants to publish on the app store, especially given the restrictions. Most developers I know in the macOS ecosystem don’t target app store for obvious reasons.
I wasn't referring to the App Store. I was referring to using the system provided APIs instead of inventing the wheel over and over again.
I agree. And that would make sense given that those type of Pro users upgrade in that time frame anyway.
Agree, especially because a person can buy the components for a dedicated, well-performing general, non-gaming PC for ~$1,000 USD
Those APIs shouldn’t change in theory, but the toolkit chain requires a full compatibility with the architecture (Just look at trying to get vscode on ARM). That’s what’ll take time.
That's not an adequate analogy. Steve suggested the Mac was the truck.
Well, the MacBook Pro, iMac Pro, and Mac Pro certainly are. Not everything those pro Macs do revolves around the Mac platform itself. A lot is web development, or Windows development, or embedded development.
Like, yeah, I wouldn't be shocked if Apple says "you know what, we don't care about that niche". But it doesn't follow that it's irrelevant.
Maybe. But again, I don't believe that's so much it has any sort of relevance for Apple in this decision.
With the exception of Windows (x86 at least) development, I don't see any reason this wouldn't be possible with an ARM based Mac.
What risk? The entire Apple power user base is an edge case.
And I do agree with the right tool for the job - and that job can't be done on ARM (today). That could change in the next decade or so, but I don't see it happening.
2021 - AMD moves to 5nm with Zen 4 (4-way hyperthreading per core). TSMC is getting good results with 3nm today. 8 threads per core has been available on the PowerPC architecture for a few years now. 4 core/16 thread baseline CPUs, anyone?
It doubled the price - I did a TCO analysis on it when it was released - $2,700 just to replace the functionality that was taken out. Added multiple points of failure with the rats nest of cabling and power bricks of uncertain quality. And then there was the whole "The D700 will die - and they can't be replaced because the ROM is on the motherboard, not the card."
It is software that is used in 3d, and I left a lot out. I have older copies of several of those (I use them to get models out of their native formats into something I can use with my current workflow) - I can run them in a VM. Won't be doing that on ARM.
Most of applications are not available on OSX now - because:
1. GPU computing - which doesn't exist in OSX (AMD is trying with the ProRender engine, but the world runs on CUDA).
2. They don't feel that they could sell enough copies to justify the investment.
3. All of that hardware is resource intensive - this kinda leaves Apple by the wayside.
Again, I can run them in a VM, because my system has the horses to do it (see sig - that is underpowered for my use - I own many poorly coded Adobe apps).
Companies aren't moving to a workflow that depends on a company that can't be trusted. Apple doesn't do roadmaps. Apple has been known to show up at WWDC and tell the audience of developers that all of the work they have spent the last year developing can't be used in the upcoming version of OSX. (Pay no attention that P.T. Barnum told them the year before that Carbon 64 is the pathway to Intel OSX.) That btw, way why Adobe and Microsoft was so late to the party on the PPC - Intel transition.
I am sure if your time isn't valuable, you could get work done on a macbook air. But workflows that need horsepower are only available if you don't mind paying 3 to 4 times the value of the parts - because someone has to pay for the development of yet another proprietary connecter, and keep someone in Taiwan making Polaris CPUs. That is why will see a new iMac Pro, btw - AMD no longer makes Vega 56 GPUs.
But if that’s for Windows, then it doesn’t really matter what Mac does, right?
The risk is: when my job consists of developing Windows x86(-64) software or working with programs that do not run on macOS, I don't buy a Mac. If I do so (for whatever reasons - design, preference of macOS, you name it), I have to be aware that I am using this Mac in a way that is not relevant to Apple when making decisions for their future.
Ad 1: I did tinker with DAZ Studio some years ago. It's sad that proprietary CUDA is so prevalent in the 3D world.
Ad2: Well, as I said, if I rely on non-Mac software for my living, I shouldn't buy a Mac. Or accept that my workaround might break every day.
Ad 3: This might be one of the reasons this switch is coming. No one knows what an Apple chip with the cooling and power capabilities of a notebook or desktop computer is capable of.
Out of all of these, only two of them currently work on a Mac. So if these are the software you are needing, looks like you need a Windows machine.