I have the M1 air and I can't imagine wanting anything smaller.
I wouldn't mind an m2 in it so it might support more than 16 gigs of ram tho!
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 Testing at Least Nine New Macs With Four Different M2 Chip Variants
- Thread starter MacRumors
- Start date
- Sort by reaction score
I wouldn't mind an m2 in it so it might support more than 16 gigs of ram tho!
Just going off of past patterns (when Intel was at their best) and how Apple has been going with recent iOS devices, it seems the most likely that they’ll support devices for the latest upgrade to software for about 5 to 7 years after that device is discontinued.
So, if Apple discontinued its last M1 product in 2023, Count somewhere between 5 to 7 years from that. So I’d say the M1 line of products should be getting the most recent versions of software till at least 2028-2030.
Plus don’t forget that even if you can’t update to the latest OS, Apple still gives the two previous generations of operating system‘s security updates for two years.
So M1 Devices shouldn’t be completely dead, as in not even getting security updates, until 2031 at the earliest, more likely 2032 or 2033.
Just going back to when Intel was at their best, the mid 2007 iMac ran all the way from tiger to El Capitan (10.4 to 10.11) and continued getting security updates until 2018 when Mojave was released.
That was 11 years, and M1 getting even longer is more likely than not
I mean not really.
There have been absolutely no rumors about anything more than the basic M2 coming out this year. Everything else is scheduled for 2023 at the earliest.
As for progress moving too quickly for consumers, I don’t think most people realize that the M1, just the basic M1, the first generation, the one in their lowest end computers… is already almost 2 years old.
It’s an absolutely brilliant processor, but it’s not exactly the newest thing in the world.
For Apple to introduce an M2 this year makes all the sense in the world
No we do not. We are hopeful, but that means nothing. Let's now get a little more specific. Correct me if I'm wrong, but this is what I have gleaned over the last little while:
M1 Mac mini - Cannot support 32 GB.* This is not strictly a marketing choice. It's a design and technical choice.
M2 Mac mini - Not sure, but looking through the leaks, I'm not optimistic it will support 32 GB either.
M1 Pro Mac mini - Can support 32 GB, but we don't know if this will be released.
M2 Pro Mac mini - Can support 32 GB, but we don't know if this will be released.
*Currently the M1 uses two chips of LPDDR4X at a density of 8 GB per chip. At the time of initial manufacture, with the design limitations and memory chip availability, it was not feasible to implement a 32 GB SKU with this architecture. I don't think 16 GB chips were available in volume at reasonable cost.
Really?
I moved from a 2016 MacBook Pro 13 inch with touchbar, to an M1 Air.
I was always impressed with that MacBook Pro, it always felt super light yet extremely solid.
With the MacBook Air, when it came to its design I was actually slightly disappointed when I first got it.
At it’s thickest point it’s thicker than my 2016 MacBook Pro was, and it’s weight never seem to be proportional to its thickness. Just holding it in the hands, it felt like it could be lighter, whereas my MacBook Pro always felt perfect.
Just for reference, the M1 MacBook Air is 2.8 pounds, and goes from 4.06 to 16.0 MM.
The current 13 inch Touch Bar MacBook Pro is 3.0 pounds and 15.4 MM.
The 12 inch MacBook was 2.03 pounds, and goes from 3.56 to 13.21 MM.
Given that the MacBook Pro has a better battery, a fan, a Touch Bar, a brighter display, and really isn’t that much heavier or thicker yet the same processor currently in the MacBook Air can fit in an iPad less than half of its thickness really tells me something.
Even if Apple couldn’t exactly match The old 12 inch MacBook, having a MacBook Air somewhere around 14 mm thick and around 2.3 to 2.5 pounds would be absolutely fantastic.
They removed the fan and put a much cooler more power efficient CPU in the last MacBook Air without barely changing the design, so it’s clear that there is a lot of room for it to get more thin and light.
Especially if they get rid of the useless wedge design. Sure it was cool… in 2008, but I’d rather have a lighter and thinner laptop all around than just one that has the bragging rights of saying “look, it’s so thin… if you look at it from the front and literally nowhere else.”
These days that’s a gimmick, it provides no actual benefit if at its thickest point the computer it’s thicker than ones without the wedge shape
Last edited:
I guess what I'm really asking is--is it now the case that if you want 32+ GB of memory you have to go Studio? Not sure they're introducing a Pro chip for the mini if the Studio is the step between Mini and Mac Pro. If I had to guess I'd say they shrink the Mini even more and give it colors like what it sounds like they are doing with the mini. That way you have to play more to get 32gb in a minimal, stationary form factor. Which i think would obviously be a shame.
If your question is really *now* then yes you have to go Studio, since no 32 GB Mac mini exists (Intel notwithstanding).
OTOH, if your question is if there will be a 32 GB Apple Silicon Mac mini in the coming year, the answer is a solid 'maybe'.
The rumors of the redesigned Air suggest something that will be lighter and more compact than the current Air, which is a case design that is around 10 years old.
I love the form factor of my 12" MacBook (2017 Core m3 16 GB). However, I must admit the new Airs are getting progressively smaller, impressively so.
BTW, I bought a 2017 Air last year, and that thing is a beast. Even though it's lighter than the 2015 MBP I got around the same time, the 2017 Air feels much more awkward than the 2015 Pro because of the Air's larger footprint. The M1 is even smaller.
But then again, the 12" MacBook still feels considerably smaller, and is a featherweight in comparison. They could stuff a 12.5" screen in its footprint too if they reduced the bezel size.
That would make an interesting product line:
12.5" MacBook
14" MacBook Air
14" MacBook Pro
16" MacBook Pro
I don't quite understand the existence of the 13" MacBook Pro.
Currently? Yes, unfortunately.
(This is perhaps the biggest reason I didn't impulse-purchase the 2021 iMac.)
However,
- there is a $800 gap between a mini (configured to be similar to the low-end Studio) and Studio, and maybe Apple wants to add a higher-end mini with the M1 Pro. But maybe they're happy with the gap.
- I wouldn't be surprised if the M2 doubles available RAM, starting at 16 GB and offering 32.
Configure a M1 Mac mini however you want, it will never match the performance of the base Mac Studio...
| M1 Mac mini | M1 Max Mac Studio | |
|---|---|---|
| CPU | 8-core | 10-core |
| Core Types | 4P/4E | 8P/2E |
| GPU | 8-core | 24-core |
| RAM | 16GB LPDDR4X | 32GB LPDDR5 |
| UMA | 68GB/s | 400GB/s |
| SSD | 512GB | 512GB |
| Ethernet | 10Gb | 10Gb |
| Cost | $1199 | $1999 |
The base M1 Max Mac Studio has twice the P-cores (but two less E-cores), three times the GPU cores, twice the RAM & nearly six times the bandwidth of a "similarly equipped" M1 Mac mini...
The only thing that is really "similarly equipped" between the two is the SSD & Ethernet...!
I think most people wouldn't bother with the 10 GigE, so the price would be $1099.
Give me a small CPU upgrade and a couple more USB ports, and as you know, I'd pay $200 more for that, so $1299. I'd even consider $1399. I personally don't need the 32 GB.
P.S. About a decade ago, I wired most of my house for Ethernet and thought hard about what I would need to support 10 GigE, cuz I thought I might be on 10 GigE by about 2020 or so.
Fast forward to the present, and I still see no reason for 10 GigE in my home. I suspect it might not be until 2030 or later, if ever, before I convert my home to 10 Gigabit.
Apple didn't "have to" in 2017 when they didn't formally announce or ship anything. They didn't "have to" in 2018 when they didn't formally announce or ship anything.
June WWDC 2022 is't particularly any significantly different from those situations. They would need a better hint than a simplistic "another day". ( In 2018 they threw out "not this year". Not overly specific either, but an adjustment on expectations. ).
Apple could easily do an announcement in October when macOS releases and there is some more Macs to do.
The more critical issue is whether it is ready to ship in next 6 months or not. If not then there is no reason to announce. I think many folks are presuming that it has to be already done a while ago and the supply chain for components is solidity producing soon. If production is ready to go in a handful of months , then yes they should give some lead time , but "WWDC 2022" isn't the "have to" factor there.
3 months ago there was talk of an iMac Pro that Apple was going to "have to" announce this Spring. And now there isn't.
There is no physical adapter/connector on an Ultra to do NUMA with another separate chip package with. It is like doing NUMA with no Infinity Fabric or QPI on. AMD/Intel. Or using a Ethernet cable when there is no Ethernet controller in a system.
The other major issue that macOS doesn't deal with with relatively high levels of "Non uniform access" of memory.
With M1 generation and macOS Apple has neither the hardware or software tools to implement this.
If someone has an old 64 bit app with OpenGL 3.0 only utiliize and throws it at a W6800X MPX module are they going to get optimal utilization out of that app on that hardware? Probably not. New GPU , not so new software ... non optimal reasons. Happens all the time on macOS , Windows, Linux , etc.
Back in December 2021 the Informatoin talked about
" ...
The 3nm chips are reported to have up to four dies with up to 40 CPU cores per chip. So they should have no trouble outperforming the upcoming M2 chips. In comparison, the M1 Pro chip uses 8 to 10 cores and the M1 Max uses 14 to 16 cores.
The M3 chip (as it’s currently called) should offer faster speeds and better battery efficiency than the M1 Pro and M1 Max chips — both of which already have impressive performance. Since the M3 reportedly won’t go into production until the end of 2022, we likely won’t see it debut in Mac devices until 2023. ... "
New Apple Macs with 'M3' chips could arrive in 2023
TSMC's pilot production of Apple 3nm chips is tipped to start late 2022 — here's what that means for Macs
That is not a "ship new Mac Pro fast" solution. Which is why some folks are tossing it aside. Because Apple "has to announce" real soon.
Waiting for TSMC N3 makes the "Max class" dies smaller (if don't crank up core counts a lot) which will make packaging four of them more tractable.
I'm expecting Mac Pro on 4nm in 2022. It wouldn't really have made sense to target 3nm for the first Apple Silicon Mac Pro. Too high risk to target 3nm that far in advance.Back in December 2021 the Informatoin talked about
" ...
The 3nm chips are reported to have up to four dies with up to 40 CPU cores per chip. So they should have no trouble outperforming the upcoming M2 chips. In comparison, the M1 Pro chip uses 8 to 10 cores and the M1 Max uses 14 to 16 cores.
The M3 chip (as it’s currently called) should offer faster speeds and better battery efficiency than the M1 Pro and M1 Max chips — both of which already have impressive performance. Since the M3 reportedly won’t go into production until the end of 2022, we likely won’t see it debut in Mac devices until 2023. ... "
New Apple Macs with 'M3' chips could arrive in 2023
TSMC's pilot production of Apple 3nm chips is tipped to start late 2022 — here's what that means for Macswww.tomsguide.com
That is not a "ship new Mac Pro fast" solution. Which is why some folks are tossing it aside. Because Apple "has to announce" real soon.
Waiting for TSMC N3 makes the "Max class" dies smaller (if don't crank up core counts a lot) which will make packaging four of them more tractable.
I'd disagree with this because I think even if the chips are technically somewhat different, right now the "M" brand is really strong. It would make more sense, IMHO, for Apple right now to ride that as hard as they can (as they have done with Pro, Max, Ultra) rather than creating a new (think "Xeon" equivalent) brand.
At some point, once the excitement around M calms down, the time may be right for a new level of excitement created by the construction of a "Xeon"-equivalent brand, but I don't think now is that time.
Yes, that gap can be a problem for sales.
There is a principle that says that you don’t want large price steps from one model to the next. Customers will chose higher configurations of the base model until they either jump up to the next product or find a price point below that they are comfortable with. If you have a large jump between the base product and the higher end product, customer will get stuck at the lower price level. They won’t be able to choose enhancements to the base that put the product above that jump price But they won’t feel that they can afford the full higher price of the higher end product.
ideally, you will have a smooth range of price points from one product to the next. Customers can self select into the configuration that most meets their needs and that is at the highest price they can afford for that. This would suggest that there should be one or more configuration within that $800 gap. Whether that is a Mini with more RAM or a higher end chip or a Studio with a lower end chip or both, we won’t know until Apple finishes fleshing out their product range.
I don’t understand why they can’t just focus on an advanced motherboard on the Mac Pro rather than building new SoC for such a niche low selling product?
Like the old Xeon Mac Pros why not just have the motherboard do all the work and allow you to drop in 1-4 ultras at your discretion? The design of the chipset of the motherboard would be the focus then. If they need to update the Mac OSX kernel, then that seems doable. Certainly that should be a technical no-brainer for them if it’s required to utilize the four ultra cores to the fullest? Maybe that will be a major feature of the next Mac OS release, enterprise server readiness.
Doesn’t this seem the most plausible step for the Mac Pro? I just can’t believe they will spend a fortune designing some frankenstein SoC for a product that makes up a few percent of the sales. But designing an awesome motherboard that supports multiple ultras and all sorts of extra enterprise features, like maybe multiple nvme modules in raid? And maybe like 16 thunderbolt ports instead of PCI slots? That sounds logical to me.
Like the old Xeon Mac Pros why not just have the motherboard do all the work and allow you to drop in 1-4 ultras at your discretion? The design of the chipset of the motherboard would be the focus then. If they need to update the Mac OSX kernel, then that seems doable. Certainly that should be a technical no-brainer for them if it’s required to utilize the four ultra cores to the fullest? Maybe that will be a major feature of the next Mac OS release, enterprise server readiness.
Doesn’t this seem the most plausible step for the Mac Pro? I just can’t believe they will spend a fortune designing some frankenstein SoC for a product that makes up a few percent of the sales. But designing an awesome motherboard that supports multiple ultras and all sorts of extra enterprise features, like maybe multiple nvme modules in raid? And maybe like 16 thunderbolt ports instead of PCI slots? That sounds logical to me.
Last edited:
More USB C/Thunderbolt ports and a higher RAM ceiling is what I'm after. 10 GigE would be nice, but I agree it's not for most home users (so not so applicable for the mini).
In March 2017 they had to hold an unprecedented press conference to admit that they needed a "modular Mac Pro". Pretty clear that it wasn't even on the drawing board at that stage. (presumably some key customers/developers they let in on their original plans were threatening to jump ship).
At WWDC in 2017 they announced the iMac Pro - which was either a cobbled-together stopgap for Pro users or was originally going to be the new Mac Pro (seems to me that the timing was right for them to have been showing (i)Mac Pro prototypes to those key customers - and getting laughed out of the house - around March).
By WWDC in 2018 they'd pretty clearly signalled via leaks and interviews that there was going to be no new Mac Pro that year.
In Spring 2022, Apple "had to" announce a replacement for the 5k iMac and iMac Pro - and they did, it's called the Mac Studio and Studio Display. They even went out of their way to say "just the Mac Pro to go" to make it clear that the 5k iMac is gone.
Problem now is that nobody has a clue what the new Mac Pro is going to be. Will it have PCIe slots? dunno. Will it take 512GB-1.5TB RAM? Dunno. Will it use discrete GPUs? Dunno. Will our current software be optimised for whatever exotic hardware it has? Dunno. Should a customer budget for new Mac Pro equipment in the contract they're currently negotiating? Dunno. Will Intel Mac Pros still be available when your lease runs out next year? Dunno... and it's been "Dunno" since Apple announced the transition to Apple Silicon 2 years ago (...only 6 months after the 2019 Mac Pro actually went on sale). As I said, these are not impulse-buy items like iPhones.
When I said "something like NUMA" I didn't mean "just use NUMA, bish bash bosh, job done". There's apparently no free physical adapter/connector on a M1 Ultra die to make it into a 4x chip either - but then we didn't even officially know about the interconnect on the M1 Max until Apple announced the ultra. Maybe there are more secrets lurking, maybe Apple can cook up something halfway between NUMA and XGrid using Thunderbolt connections. Maybe Apple could make a variant of the M1 Max by hanging an inter-chip interface off the existing interconnect. Maybe the M2 Max/Ultra will feature an interconnect (a M2 Ultra Mac Pro would sit uneasily with the Studio Ultra, but a Mac Pro with multiple M2 Ultras would be distinct).
The point I'm trying to make is that the hardest, most expensive option would be to produce a completely new die - beyond M1 Max - in tiny quantities just for the Mac Pro. It's not impossible, but if you rule that out, the alternatives are either stitching more and more Mx Max dies together (still fairly expensive) or going for some multi-processor solution (much cheaper hardware-wise but shifts the problems to software).
Or maybe the Mac Pro will just be a rackmount Studio Ultra and they'll get third parties to make rackmount Thunderbolt PCIe enclosures and storage units to go with it.
That comes with multiple problems.
RAM: each of those Ultras has its own RAM. Does each process only get access to one particular SoC's RAM? Do you get rid of the per-SoC RAM and instead have a shared pool? (But then: much worse latency!) And even if you do that, no upgradeability, and even with four Ultras, you'd only get a third of the RAM as with the current Pro.
GPU: each Ultra would appear as its own GPU device in Metal.
And we haven't even gotten to things like: how do you synchronize the PCI bus?
No, I think they'll need a significantly different design for this to make sense.
I purchased my M1 Max when it became available to buy and took delivery on it when Apple said I would. Today given what is going on in China and the rumors of what may go on in China Apple needs to find fabrication in other parts of the world for their products. I did take a chance with my M1 Macbook Pro 14' but I am totally happy with it. I am very glad to read about Apple moving ahead with progress of newer progress and would never expect Apple to slow down progress of technological development just because I own an older technology. It was very much what I expected when I purchased my Macbook Pro 14"! I am sure that anyone who buys the current generation of technology will be satisfied