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.
Anyone else used to be very pro-ARM but now having second thoughts?
- Thread starter Spungoflex
- Start date
- Sort by reaction score
I understand all of that. But there are some dev
You think an iOS app would be that slow emulated on Intel?
Well, for one, I think it is more efficient to translate x86-64 to ARN then the other way around. ARM has more registers and can load/store multiple registers at one, so I would expect a translation of an optimized ARM64 to be an overly verbose, suboptimal x86. Also, Apples ARM CPUs are going to be faster than a Intel ones. You can reasonably take some performance penalty to emulate a slower CPU using a faster one, but not the other way around. I feel that this kind of strategy would make Mac a bad experience for the users. Just look at the disaster Microsoft caused with 64-bit adoption.
Another issue us that ARM Macs will have access to APIs that simply don’t exist on the Intel side. You can forget about running any app that uses Apple-specific Metal features for example.
I jjst wish that they would give us more details on dates. I am on the verge of really needing a new desktop but being as the nonARM Macs will only have support for 4 more years id rather wait if i can. But why are they even releasing another onARM updated IMac in 3Q if they want to release the ARM based one late 4Q?
GTA III was ported to ARM a decade ago. You can buy it for a few bucks on the iPad app store, and unless they deliberately remove it, the game will be available the ARM Mac App Store along with pretty much the entire iPhone/iPad game catalog.
And just because intel is going to be "supported by Apple" doesn't mean it's going to be a first class citizen. You won't be able to play GTA III on it for example. Because they won't bother to emulate ARM apps on Intel - even though it could easily be done (especially for old/low tech games like GTA III).
Rather than wondering if ARM will work out for Apple, I am more interested as to what Microsoft (and subsequently Intel/AMD) will do going forward.
We know that more than a decade ago Microsoft was already thinking about creating a new OS to do away with legacy issues on Windows which was holding it back in terms of innovation - Apple moving to ARM will almost certainly push Microsoft to work harder on a transition and try find a bridge gap solution to keep their “might still be on Windows XP” business clients happy.
We know that more than a decade ago Microsoft was already thinking about creating a new OS to do away with legacy issues on Windows which was holding it back in terms of innovation - Apple moving to ARM will almost certainly push Microsoft to work harder on a transition and try find a bridge gap solution to keep their “might still be on Windows XP” business clients happy.
Honestly, they can do the same as they did with 16-bit code before support was deprecated. Instead of being able to directly access the operating system, just keep the code isolated under a virtual compatibility layer. It is the approach they are taking with Windows 10 X.
What's best about this approach is that they don't even need to to distribute the compatibility layer straight away. They can keep the OS small, and if the user needs legacy code, just allow them to download the legacy compatibility layer as needed. This can even be done offline is connection is an issue.
It’ll be an interesting space to watch.
In 6-12 months, we might be having Apple laptops that walk over Windows laptops performance wise, heat/noise wise and form factor as well (can keep same performance in smaller package) - and none of the Windows manufacturers will be able to compete, at least on the current chips they use.
Telling users “but ours can run x y z which you can’t on ARM” although true, will not be applicable to the vast masses and they might find they are bleeding business as a result over the next few years.
This is why I’m fairly certain they will have to take some radical decisive action soon. I don’t even know if merely switching to AMD CPU’s will solve their Intel hell issues.
I am still all in. Why? At WWDC, they said that MS Office would run natively on Apple Silicon - that is my main use.
My daily driver is an rMBP 13“ early 2015. Tried a 10th gen MBP some weeks ago which was a disappointment, getting hot and loud too often. Intel doesn‘t move forward. I will stick with my loved machine, just got a new battery before it is going vintage and put a bigger SSD into it. So I can wait and see whether Apple delivers, but I am optimistic. Typing this on a iPad Pro 1st gen, great device, the same technology in a fanless MBA with the macOS UI and much more power than the MB 12“ - done deal
My concern with Microsoft and Adobe is, will it run natively or will it run natively well? Will they focus a lot of their attention on performance and optimization? Adobe already performs poorly compared to Affinity Photo/Designer on macOS. Adobe is better, far far better on Windows. Same with Office.
This is why I am holding off for at least a couple of years after the transition. So I am looking at ~4 years before I get one of the new systems.
[automerge]1594313688[/automerge]
There is NO chance Apple will axe the Mac Pro. Apple got so many complaints about the 2013 Mac Pro that they publicly apologized for the situation and said they would make things right again. If this happens again, I think it will leave a black eye/dark cloud over the entire Mac lineup with the pro community for a very very long time. I for one will be done if this happens. It will be a second time Apple ignored the pro market to a degree that is ridiculous.
Apple cannot do this unless they want to get out of the Pro market. So I do not see dropping the Mac Pro would be a consideration.
There is a good chance we don’t have AMD graphics at all for ARM Macs based on some documentation (explained better in the video below)
I didn't say anything about ditching the Mac Pro, I was talking about the iMac Pro...
YES!!! I'll gladly trade in the ability to play RDR2, Fallen Order, Witcher 3 in bootcamp on my MBP 16 for Candy Crush and some Animal Crossing!! /s
Seriously though, Mac gaming selection was pathetic during the PPC era and we're definitely going back there with AAA titles never being ported or about 2-3 years later.
I'll just continue to use my MBP 16 for now and later get a 2nd/3rd gen ARM and might just have to carry a Razer 13 for gaming on the road....
I will be fine as long as WoW is transitioned over, I can get a PlayStation if I want to play the latest GTA.
Although I did play WoW for quite some time on my unibody 13" MacBook w/ Nvidia 9400M GPU which used system RAM. PuGged Dragonsoul for 26 weeks to get my Legendary Daggers! Warlords of Draenor was too much for it though, so when Legion came out I threw together a spare parts PC. Now I am rocking a 3900X & GTX 660 Ti (2GB RAM, YO!), waiting for RDNA2 & then Ryzentosh time!
And then I wait for my shiny new Apple Silicon arm64-based Mac Pro Cube! ;^p
With virtualization, Microsoft wouldn't really have to allow it; VMWare or Parallels (the latter of which I believe is already working on it) just needs to have an x86 emulation shell sitting on top of the OS. It will be potentially a lot slower, like the old days when people ran Windows VM's on G4's.
I think Kuo has missed the mark. His two predictions make no business sense in light of information we have. Knowing my luck the 16" MBP will be first with ARM.
If Apple starts churning out faster, cooler, thinner, sexier machines with better battery life, PC manufacturers will want to try ARM. Hard to say who will make the CPUs for them but Microsoft will want to cater and probably get Windows 10 or 11 running with 64-bit support on ARM. I imagine they'll license it for virtualisation on Mac too to make sure they are monetising it as best they can.
I can empathize with your feelings here.
On paper, Apple transitioning the Mac to their own ARM SoCs makes too much sense. Assuming developers play along and transition their apps over, it will be a fantastic move for the future of the platform, especially given their desires to provide parity between macOS and their other platforms. If you are an avid macOS user, then the potential is pretty great!
However, whatever PC gaming was able to be done on a Mac (either natively in macOS or via Boot Camp) takes a hit with this move. And the freedom to run other operating systems on your Mac (either via dual-booting or via VMware Fusion or Parallels Desktop) takes a hit given that there are WAY more x86 operating systems that you can run out there than there are ARM ones.
Seeing as these are two very important things for me as a computer user (let alone a Mac user), I'm not going to wait for ARM. I'm going to get an Intel Mac in the not-too distant future and that'll be what I roll with. Macs will Apple Silicon will be waiting for me. Suffice it to say, if your main second thoughts revolve around these functions, then I definitely empathize. The Intel Mac era was a great era that produced the best of the Macintosh, by far. We know little about what's coming.
I'd say that, if you can afford it, get a really great Intel Mac today, and keep an open mind for the future. Certainly, as far as macOS and compatible apps are concerned, we've got some pretty great performance coming up in the new era.
Yes and no. Currently, the server variant for Ubuntu has an ARM port. The standard non-server version does not. ARM is still up and coming as a desktop architecture thanks to the Raspberry Pi (remember, it's much more common as a mobile architecture). Your mileage will vary, but for now, assume that selection is still somewhat limited. This may be why Apple's current stance on direct booting a non-macOS OS is what it currently is.
32-bit x86 Windows software runs fine on Windows 10 for ARM64. The emulation layer is slow, but that's because, up until now, the SoCs used on ARM64 PCs have been slow. You're only unable to run 64-bit x86 software. However, Windows software that is exclusively 64-bit x86 is extremely rare given that Microsoft hasn't put the gun to developers' heads about it the way Apple did with Catalina.
Unlike Kelis' Milkshake, having the better (more efficient, faster, etc.) platform doesn't necessarily bring all the boys to the yard. This is why tvOS hasn't taken off as a console gaming platform despite both supported Apple TV models having adequate graphics performance for console-level gaming. Apple's SoC's are powerful and efficient, but that doesn't necessarily equate support from developers that have spent most of their time developing for more traditional computing/graphics architectures (despite their relative inefficiency).
I bought the first Intel 20" iMac back in 2006. It was fun being an early adopter to that switch, despite not having many of my pro apps function for a bit or many of my games work well (or at all). And I'm grateful that I got the 20" iMac rather than the first 15" MacBook Pro.
I think this time, I'll be content to buy a final gen Intel Mac and watch (and support) others buying early, so that when I make the jump a few years down the line, I'll know what I'm getting into and be more ready to get into it.
I would imagine that a 2018 15" MacBook Pro would remain supported for macOS releases at least through 2026, if not a hair past that point (and don't forget that you still get security patches on your last supported OS for two years after that point). These days (as opposed to the days at the start of the Intel switch), Apple drops support for machines it either can't procure updated versions of every driver for or that don't support a feature they deem critical to that release of macOS (i.e. Mojave dropping support for Macs with GPUs/IGPs that don't support Metal). Apple will inevitably have incentive to stop updating the Intel codebase, just as they did with the PowerPC codebase with Snow Leopard, but those days are a long way off (and Apple has many many more active Intel users now than they did PowerPC users at this point in that transition).
7 years is a little conservative, given that 2012 Macs are only now getting left out of the next major OS release some 8 years later. Mind you the predecessors for most of those Macs had 7 years (though some of their predecessors lost support on the same day). Our mileage will totally vary in terms of when a Mac loses support. It's safe to say that systems based on 9th and 10th Gen Intel will be supported for a good while. 8th Gen too. The only exception that might be worth worrying about is the current iMacs. If Apple deems a feature important to the platform that requires a Mac with a Secure Enclave in the same fashion that they deemed Metal support on Macs to be critical, we might see the dropping of support of every Mac that doesn't have one (which would mean that 9th Gen based Intel 27" iMacs would lose support sooner than some 8th Gen based Intel Macs in that hypothetical scenario).
They may not hold and more monetary value for being Intel Macs. But certainly, you will be able to run newer versions of Windows 10 after your Apple has stopped supporting your Mac for new macOS releases. Case in point: 2012 Macs will STILL be able to run a current version of Windows 10 after Big Sur leaves them out.
This is not the PowerPC to Intel transition. Apple has many more Intel Mac users heading into this transition than they ever did PowerPC users headed into the Intel transition. Developers will be foolish to drop support for Intel prematurely. I do agree that it will inevitably happen. But it won't happen that fast with most key developers. Where you will likely see this will be with Mac games that were originally ported over from Windows (but never fully received the Metal make-over). You'll also likely see it with software that doesn't get regular rolling updates. I would also guess that software that made the move to 64-bit begrudgingly so that it could still run in Catalina is also in danger. But then again, for much of those titles, you simply might not see an ARM release. Porting x86 Windows code to x86 macOS isn't as big of a challenge as porting x86 Windows code to macOS on another architecture. This is why, following the switch to Intel, we saw a surge in new Mac software (games especially included). It's much more likely that you're going to see some Mac developers leave the platform due to having mainly been there for x86 parity.
Again, I think that you're going to continue to see Universal Binaries for as long as Apple is releasing Intel-supported releases of macOS. As soon as Apple drops support for Intel in future macOS releases, that'll be when Universal Binaries follow suit.
Mac gaming had a surge in 2010-2012 with Steam on the Mac. My guess is that Valve is not going to be keen on supporting ARM in any way (they only have two of their games with 64-bit Intel at this point!). Considering how few Mac games from Aspyr and Feral Interactive survived Catalina's Culling(tm), this could be indicative of the end of triple-A titles from PC also making their way to Mac.
I do have a feeling that Blizzard will update their non-Overwatch games for the new architecture. I do, however, think that the Apple Silicon switch means that it will be less likely that we will see an Overwatch Mac release, however.
Emulation and Virtualization are not the same thing. An x86 operating system running in VM on an ARM processor requires emulating that architecture. And, like you said, there will be a speed hit. The person you were replying to was talking about virtualizing the ARM64 version of Windows 10 which has licensing constraints (imposed by Microsoft, for the time being) that the x86 and x86-64 versions of Windows 10 do not. Virtualizing the ARM64 version of Windows 10, while not being helpful for those wanting to run a 64-bit x86 Windows app, would still be faster than emulating the x86 architecture for an entire system. VMware has all but stated that a would-be version of Fusion for Mac for ARM wouldn't run x86 VMs. It's very likely that Parallels will have a similar policy and/or such a policy might even be dictated by Apple.
The 13" MacBook Pro is looking like it will be the first with ARM, not the 16". This isn't a matter of making "business sense" or not. Apple has the capability to put out a machine with their SoCs that handily outperforms any and all Intel-based 13" MacBook Pros TODAY. The same isn't necessarily a given for all Intel-based 16" MacBook Pros TODAY. Apple might be able to best the 6-core i7 models TODAY, but the 8-core i9 models aren't necessarily a given. And Apple is not about to release an Apple Silicon 16" MacBook Pro replacement that doesn't outperform every 16" and 15" MacBook Pro that has ever existed. THAT would make no business sense.
As for Windows 10 on ARM64, I agree that Apple's hardware is a good reason for Microsoft to consider partnering with them to either get an Apple Silicon Mac dual-booting solution in order (a la Boot Camp for Intel Macs) or to at least get Windows 10 on ARM64 virtualized on Apple Silicon Macs. Macs benefit from supporting Windows and Windows 10 on ARM64 needs competent hardware to run on to demonstrate to Microsoft, consumers, and Windows application developers that it's not a lost cause.
No ARM Macs for me as I already have a locked down iPad and iPhone which I can't downgrade to older FW. Why would I want to do that with my Macs too?
Link to where the MacOS will be locked down please.
Looking like the first is based on that one report from one analyst, its just everyone is reprinting it. He also said the iMac would be the first and we already know there is an Intel iMac coming before an ARM one. Is Apple going to release an iMac and then supersede it within 6 months? I know they did that last time but I can't see them doing it again. The 13" is more likely than an iMac as its been out longer but it's performance is not that far behind the base 16" as it is. An ARM one will likely beat the 16" at a lot of tasks, particularly single threaded. If it doesn't, ARM Macs will be considered underwhelming. Apple needs the first one to make a real splash and I think it will. The flagship MBP was first last time, its more due an update and its price point means it won't cannibalise other models much. Plus I just bought a 16" so its bound to be replaced imminently.
I get the feeling Apple are no longer interested inane benefit they got supporting Windows. I consider that an indicator they expect to grab some significant market share with Apple Silicon Macs (there's another thread talking about this), but if ASM are as compelling as I expect them to be, other makers are going to start looking beyond Intel and they will need Microsoft to support whatever Qualcomm or whoever manage to copy from Apple's new systems. Microsoft will happily follow suit as they have shifted substantially from being a company that sells operating systems to one that sells services. They are going to want to maintain the position of Office and 365 etc and that is likely to require Windows on ARM sooner or later. That will shift the ball back into Apple's court but I expect them to be fine with running Windows ARM VMs and that will keep quite a few developers happy I suspect.
I've been pro-ARM machines since Newton was a brand of PDA. They were exceptionally expensive in those days.
Having gone from Motorola 68040 to PowerPC 604e to Intel Core i7, I'm ready for something much more efficient. The CPU temperature is often 100 degrees C when I'm working on video or photos. That's really not necessary.
If Apple can provide enough processing power and have the software I use available, I'm onboard with it. It took a while to get Mac OS X off the ground. They must have an almost seamless transition.
Having gone from Motorola 68040 to PowerPC 604e to Intel Core i7, I'm ready for something much more efficient. The CPU temperature is often 100 degrees C when I'm working on video or photos. That's really not necessary.
If Apple can provide enough processing power and have the software I use available, I'm onboard with it. It took a while to get Mac OS X off the ground. They must have an almost seamless transition.
Good video..
I reckon Apple 's goal is to do everything themselves.. That way you can have total control of when stuff comes out and minimize delays and production.
Just like Google and Facebook datawars (Who can beat the other) the same will happen here..
All this is clearly driven by the fact they wanna give users want they want, even if that means abandoning third part graphics and cpu's...
On the negative, that means no more Bootcamp, (no x64 architecture), and i wouldn't wanna run VWware via Rosetta otherwise it would be far too slow,, it must be hardware based VT, or don't do it.
One question that wasn't answered in the video if dedicated graphics possibly in MBP would be removable. My thinking is "no" simply because its a laptop... and like like SSD's being soldered on newer Mac's the argument is their more reliable, and otherwise can sneak out of their sockets. (e.g during reassembly, i you don't put the SSD all the way in, or off center, it can cause damage)
And as Apple always users the reason "to conserve space" ... well. that's debatable.
The most likely thing is that the GPU will be part of the SoC. At any rate, it certainly won’t be removable. Very few laptops offer removable GPU in the first place and I don’t think that any Apple laptop ever did.
[automerge]1594368115[/automerge]
Quite on contrary. A boost in graphics performance across the entire range, and most importantly - predictable performance - would make games like Overwatch on Mac possible. Until now, porting competitive shooters didn’t make much sense because only few Macs would have the performance to run them, and even then your have to deal with random lag spikes due to buggy drivers.
Mind, I’m only taking about technical feasibility here, not whether it makes financial sense fir Blizzard to do so.
Last edited:
This makes sense.
If I was a triple A developer, my decision to make a game on macOS would be based on
1. How big is the target audience
2. Of that audience, how many can actually run my game and enjoy it
I think 2. is currently a problem, as it usually requires a high end mac, possibly eGPU if you don’t have the 16” and on top of that bootcamp. There are a lot of barriers to gaming as things stand for Macs, but if in the future they are actually really powerful machines, game developers will no longer have to worry about 2. and Mac users can just install and play without fiddling with bootcamp and other nefarious settings.
And 3. How technically feasible it is and what is the development effort
ARM Macs will solve 2. and 3. If the lowest ARM Mac GPU is just 50% faster than the iPad Pro, that would already make it better than at least 80% of Windows laptops. As to point 3., if all Apple Macs have the same GPU (just different number of cores etc.), then you don’t need to test as much. This actually makes the development cost lower.