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.
What will be your first ARM MAC?
- Thread starter UBS28
- Start date
- Sort by reaction score
By 16-64 core I assumed performance cores. Even then they likely won't go for many efficiency cores, I don't think, so the number will be similar. And I just can't see a Mac mini with a config higher than a 16-core count.
This is a great thread! I think @arn could consider featuring it on the front page.
As for me, my first ARM mac will be either a mini (although I will probably keep my current one because of Bootcamp), or a new MacBook Pro if they aren’t super expensive (yeah, even more) or a 12” MacBook if they are not severely underpowered.
However, because I plan to make a long term purchase, I probably wait until the 3nm SoC in 2022 or 2023, why? There are several reasons, being longevity the main one, also I don’t need a new mac yet, I can wait a bit more. Other reasons are: Sometimes the first batch, the first iteration of a new model may have some sort of manufacturing issue. Staingate, flex cable of the screen on MacBooks, Butterfly keyboard, you name it. So I prefer to wait until this are ironed out.
Another reason to wait is to give time for the transition to complete, give time to developers to adapt their software to the Mac platform, and to have a better scope of how Intel apps behave on the new architecture thanks to Rosetta. Also by this time we’ll be able to see if the memory needs (RAM) stay the same, go down, or go up needing more memory on the newer macs. The power consumption and graphic performance will also be interesting to see, as well as the different categories on the MacBook side (12”, Air, Pro...). I think by 2022 or 2023 we will have a better scope, and much more powerful or efficient macs thanks to the 3nm process.
So yeah, sooner or later I will jump to ARM on all my devices (Except my PS5) but I’ll try to wait until the technology has evolved a bit. Second gen devices are usually the best in terms of lifespan and reliability. Yeah, I’ve always bought the ”s” iPhone iteration, the iPad 2 was one of the most versatile devices for its time, and the iPad Air 2 has been one of the best supported iPads ever (even with iOS 14!). Not a fan of first iteration of any model, that’s why I’m not sure what to do regarding the iPhone 12, because it will be the last one with a power and I/O connector (I need that connector!), but at the same time the first one of a complete redesign, and the first with 5G... I don’t know
As for me, my first ARM mac will be either a mini (although I will probably keep my current one because of Bootcamp), or a new MacBook Pro if they aren’t super expensive (yeah, even more) or a 12” MacBook if they are not severely underpowered.
However, because I plan to make a long term purchase, I probably wait until the 3nm SoC in 2022 or 2023, why? There are several reasons, being longevity the main one, also I don’t need a new mac yet, I can wait a bit more. Other reasons are: Sometimes the first batch, the first iteration of a new model may have some sort of manufacturing issue. Staingate, flex cable of the screen on MacBooks, Butterfly keyboard, you name it. So I prefer to wait until this are ironed out.
Another reason to wait is to give time for the transition to complete, give time to developers to adapt their software to the Mac platform, and to have a better scope of how Intel apps behave on the new architecture thanks to Rosetta. Also by this time we’ll be able to see if the memory needs (RAM) stay the same, go down, or go up needing more memory on the newer macs. The power consumption and graphic performance will also be interesting to see, as well as the different categories on the MacBook side (12”, Air, Pro...). I think by 2022 or 2023 we will have a better scope, and much more powerful or efficient macs thanks to the 3nm process.
So yeah, sooner or later I will jump to ARM on all my devices (Except my PS5) but I’ll try to wait until the technology has evolved a bit. Second gen devices are usually the best in terms of lifespan and reliability. Yeah, I’ve always bought the ”s” iPhone iteration, the iPad 2 was one of the most versatile devices for its time, and the iPad Air 2 has been one of the best supported iPads ever (even with iOS 14!). Not a fan of first iteration of any model, that’s why I’m not sure what to do regarding the iPhone 12, because it will be the last one with a power and I/O connector (I need that connector!), but at the same time the first one of a complete redesign, and the first with 5G... I don’t know
I'm putting this down in part so I can come back in a couple years and see how accurate my own predictions were:
Unless there's an unexpected upgrade of my work computer, mine will almost unquestionably be a 16" (or whatever is in that range) MBP to replace my semi-loaded 2018 15", and will be purchased either in about a year when my extended warranty is up or when the first relatively-high-spec model in that size is available, whichever comes later.
I owned a first-gen 17" Intel MBP and was perfectly happy with it throughout its life, so have no particular fears about 1st gen Apple tech, and I'm frankly kind of excited to get to play around with the first generation if I can, although in reality I kind of expect that there won't be a noticeable difference in the early stages apart from battery life.
On the other hand, I just ordered a nearly-top-of-the-line 27" Intel iMac. My 6-year-old existing one isn't doing so hot, and while part of me wants to get it to limp along until a 27" ARM iMac is available, that's looking like it's probably not in the cards for a little while, so I might lose that bet and get stuck emergency-buying. Additionally, my (only slightly educated) calculation is that the performance of a 10-core i9 should be at least comparable to an A14X (or whatever) in the first gen, so this gives me ~5 years of a Boot Camp hedge for very occasional Windows-only games and work stuff that I'm skeptical I'll be able to get done with virtualization, at least not for a while.
I could lose that bet and it turns out the 27" models come out faster than I expect, and/or are monstrously fast, and/or virtualized x86 works way better than I think it will. In which case, oh well, it's still a very nice computer and will work just fine.
I'm also less worried about x86 abandonment for OS upgrades than I would be since the desktop is primarily my wife's computer that I use less frequently when I need serious CPU and GPU grunt, and I'm not that aggressive about OS updates on it, so if it gets abandoned after say 3 years I won't feel bad getting 1 or 2 more years out of it before replacing it.
Unless there's an unexpected upgrade of my work computer, mine will almost unquestionably be a 16" (or whatever is in that range) MBP to replace my semi-loaded 2018 15", and will be purchased either in about a year when my extended warranty is up or when the first relatively-high-spec model in that size is available, whichever comes later.
I owned a first-gen 17" Intel MBP and was perfectly happy with it throughout its life, so have no particular fears about 1st gen Apple tech, and I'm frankly kind of excited to get to play around with the first generation if I can, although in reality I kind of expect that there won't be a noticeable difference in the early stages apart from battery life.
On the other hand, I just ordered a nearly-top-of-the-line 27" Intel iMac. My 6-year-old existing one isn't doing so hot, and while part of me wants to get it to limp along until a 27" ARM iMac is available, that's looking like it's probably not in the cards for a little while, so I might lose that bet and get stuck emergency-buying. Additionally, my (only slightly educated) calculation is that the performance of a 10-core i9 should be at least comparable to an A14X (or whatever) in the first gen, so this gives me ~5 years of a Boot Camp hedge for very occasional Windows-only games and work stuff that I'm skeptical I'll be able to get done with virtualization, at least not for a while.
I could lose that bet and it turns out the 27" models come out faster than I expect, and/or are monstrously fast, and/or virtualized x86 works way better than I think it will. In which case, oh well, it's still a very nice computer and will work just fine.
I'm also less worried about x86 abandonment for OS upgrades than I would be since the desktop is primarily my wife's computer that I use less frequently when I need serious CPU and GPU grunt, and I'm not that aggressive about OS updates on it, so if it gets abandoned after say 3 years I won't feel bad getting 1 or 2 more years out of it before replacing it.
13" MacBook Pro (or 14" if that changeover happens when the rumor mill suggests it will). Though, I won't be buying for at least the first two years. I was first to the last processor architecture transition. It was fun. But I don't need to be first in line for this one (especially as I'll be consulting for people that ARE going to be first in line for this one). I also have need for Intel/x86 specific features of Intel Macs for the time being. I will, however, watch the Appel Silicon Mac advent very closely.
I hope there's a MB/MBP size laptop introduced from the get go.
I'm also assuming Apple will have perf as fast or faster than the new Tiger Lake chips coming out in the Dell XPS 13:
i7-1185G7
I'm also assuming Apple will have perf as fast or faster than the new Tiger Lake chips coming out in the Dell XPS 13:
i7-1185G7
That's quite an impressive Geekbench score for a 4-core CPU. My 2019 8-core MBP16 is worse in single-core and only gets about 7000 in multi-core. I would be pleased if a new ASi MacBook can match these levels, or at least gets close, while offering longer battery life, and running cooler & quieter.
I think Intel has done quite well with Tiger Lake, and it is certainly a strong competitor to whatever Apple comes out with this year. My guess is that the first Apple Silicon Mac won't greatly surpass this, or will even do slightly worse, but will offer other advantages, such as lower power consumption (very important for portable devices), and optimization for certain workloads (e.g. video & photo processing).
Last edited:
Impressive indeed. The closest thing Apple has to it is the A13's >1300 scores. I can see them matching this with A14-based chips (even assuming the derived estimates from the iPad Air 4 are accurate and also what we'll see in the X versions) but I'd be surprised if it's better by much, if not worse.
Likely a high end Mac Mini, with an eGPU and a ~34” 21:9 monitor. But if it turns out the Mac Mini is quite slow CPU wise, I will likely go Windows.
13 or 14 inch Macbook / Air / Pro whatever will have performance similar or better than Macbook Pro 13" 2018 with 16 GB RAM and 512 GB SSD. Also I will wait for info about all apps that I use if they work in Rosetta 2 or are AS native. Currently I'm on Late 2013 Macbook Pro 13" so I hope it will endure a bit more as I see no sense to buy Intel Macbook now.
^This.
I have a late-2019 MBP16, so have no immediate need for a new computer, but would like more lightweight Mac when air travel becomes possible again. I bought the MBP16 for video editing, so I'm hoping that a new 13-14" MacBook will offer better performance than current MBP13s.
If the ever fabled 14" MBP happens, that. If it happens, preferably with 32GB RAM minimum as well.
Cost won't be an issue, as i'll just order it through work to slap our prearranged 30% discount on it (en masse volume ordering + Higher Ed institution).
Cost won't be an issue, as i'll just order it through work to slap our prearranged 30% discount on it (en masse volume ordering + Higher Ed institution).
Nobody answered this, and while I'm quite sure there's much better-informed speculation in other threads by other people, the short answer is very likely no, at least yet.
I'm old enough to remember when having multiple cores meant two physical G5 CPUs in a Power Mac G5 (or a whopping four physical PPC 604e CPUs in a Daystar Genesis clone!), so it's occurred to me that as a solution for high-spec iMac, or iMac Pro, or Mac Pro CPUs Apple could theoretically just stick multiple physical CPUs in like modern Intel or AMD based servers, rather than building custom chips with more cores. Thermally, it's entirely feasible--extrapolating very heavily and speculating some, a hypothetical 8-performance-core A14Z would probably draw only around 40W while being performance-competitive with a 125W i9.
But it's not as simple as just building a motherboard with sockets for two (or more) physical CPUs--the CPUs must be designed for paralleling (on x86, only the server-grade chips have this capability, which is why there are no multiprocessor Core i[anything] motherboards), and there's a host of design considerations regarding interconnects and sharing (or more often not sharing) memory and other resources that a single CPU with many cores handles entirely internally. That also adds overhead, so the performance increase isn't linear. There's a reason that Intel and AMD are cranking out CPUs with dozens of cores instead of just putting a bunch of CPUs onto a single motherboard.
That said, Apple is going to have to address this within the next couple of years or so, so the question is whether there's enough sales volume on higher-end iMacs and Mac Pros for them to custom-design and fab many-core A-series chips just for these products, Apple will just force everyone live with A15X (or whatever) performance, or their long-term solution is going to be baking multiprocessor capability into the A15Z (or whatever), set up MacOS to take advantage of it, and return to the multiprocessor days. If anyone is in a position to make that happen, both in terms of resources and the vertical development team, it's Apple.
We'll see in a couple of years!
Thanks for answering my doubt/ query. Yeah, I was referring to Powermac G5, which had Dual Processor setup, also the laptops and specially the desktops do not have the issue of heat dissipation or power consumption like the iPad or iPhone. Both laptop and desktop are performance based solutions.
Arm based processors have to deliver better performance than the existing i9 or xeon based processor on Mac Pro. Also Apple OS should be able to take maximum performance out of arm based processor setup.
Maybe in a few years we would know how the option of switching to arm based processor is. This is the third time when Apple has changed their architecture (1. switching to PowerPC, 2. switching to Intel and 3. now switching to arm)
Given that Oracle are adopting fully ARM hardware for their enterprise cloud services from next year courtesy of Ampere ARM based CPU's, so it's not just a theory any more - HPC via ARM is no longer just a pipe dream. Also, because of that, Apple will struggle with trying to say the architecture prevents expanding to that scale. So, with a little bit of extrapolation it should be fairly safe to say that ARM could quite easily replace a current Xeon W or Xeon Platinum, because that is exactly what is happening for Oracle.
When did Apple say anything at all about the scale of either ARM or Apple Silicon being scaled to cloud servers? For that matter, when did Apple say anything about cloud servers?
for sure iMacs... i have a 2015 27 inch iMac and a 217 21.5 inch iMac that both are getting long in the tooth. I'll replace the 27 inch iMac first with Apple Silicon so long as it seems they have a processor targeting performance in it and not a quick low end model with a better Apple processor if I wait a year or something. The 21.5 inch iMac is less in need of an update because that is our family kitchen area "general computer" that is used mostly for internet type stuff that isn't too taxing.
Technically they don't have to do anything. Pro users would certainly like them to, and they probably will, but Apple as a company, and even the Mac platform, will survive even if the Mac Pro ends up with the same CPU as a high-end iMac, or if the top of the line consumer desktop gear turns out to be slower than equivalent Intel products, or if the product line is dropped entirely. Apple could even just decide to make only laptops, and anyone who wants a non-docked desktop, too bad.
They will lose some users, they will lose some clout, but these things are always technically an option, and depending on sales volume, use patterns, and business calculations, are at least possible even if they're not likely.
They didn't? The point MisterMe was making is that major companies other than Apple are already planning on ARM for big-iron servers, so high-performance ARM-based computers are not at all theoretical anymore, even if Apple wasn't getting into desktop ARM (and, presumably, high-end-pro desktop ARM eventually).
I don't think they did, but think about it. Part of the reason Intel is successful in the cloud world is because devs were running the same Intel hardware on their computers. Now Intel has gotten lazy, AWS has shown viability of Graviton and Apple will encourage devs to just develop on ARM64 compatible hardware up and down the stack.
But most importantly, that cloud that AWS dominates also generates lots and lots of cash. Apple wants their share of that cash. They can start by expanding iCloudStorage... next make XCode work on an iPad attached to iCloud for the heavy lifting... Probably be pretty seamless over 5G. Then box out AWS/Azure/GCS. It's the ultimate Sherlock.
Reasoned explanation of the fact that Apple Silicon replacements for the Mac Pro being fully feasible, despite deniers trying to deny such a possibility. The concept is called a giving a contextual example.