By the way, talking about memory solutions, here is a little patent someone linked on RWT:
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.
Rumored pro chip
- Thread starter cmaier
- Start date
- Sort by reaction score
Multi-CPU, in this case SoCs, would make it extremely complicated for the OS and applications to full utilize. The reason why servers have multiple CPUs is that cloud companies sell resources by the core and each core can be used by different customers at the same time.
Your explanation suggests that very heavily.
Multi-Processor servers are usually setup to run multiple unrelated (from the CPU's point of view) services in parallel (not always, but usually). The losses that multi-core aware desktop programs would suffer (relative to a higher core single processor) aren't really relevant if the threads they're running are from unrelated programs.
ONE such thing, is virtualisation - virtual machines aka "VMs". But it's far from the only use, and "Cloud companies" are far from the only place where virtualisation is used.
The distinction between "M1x" and "M2" - if such products ever appear - will be that one says "M1x" on the lid (and in Apple's CGI impressions of the mainboard) and the other says "M2" - which will be decided by Apple's marketing people at a late stage of development. (If Apple have published a document somewhere specifying something like "the number specifies the core version, the suffix means extra cores" then I stand corrected - although even Apple can change their mind).
I'm not sure that people discussing rumours here even agree on what exactly M1x and M2 mean (apart from the usual "exactly what I intend it to mean") - maybe, but if you get to the point of "the MacBook Air can't have that because that would be M2 and the Air has to be M1x" the wheels start to fall off. It's not inconceivable that, by the time next November, rolls around, Apple will have some significant new technical improvement to put into the Air and iPad SoCs that deserves a name that suggests more than "warmed-over M1".
Personally, I suspect that Apple will have to find a new scheme for "marketing names" or, a few years down the line, they'll be in the position where the MacBook Air has an M4 (because it uses a new low-power core design) and the iMac Pro "only" has a M3z (which nevertheless hoses the M4) - which doesn't matter one jot for technical purposes but won't play well with marketing. I'd have thought they need to develop some sort of separate "ultra-mobile", "general", "workstation" branding that can stay stable over multiple generations. Doesn't have to be quite as cynical as Intel's i3/i5/i7/i9 which didn't mean anything consistent apart from signifying the price-point of the end product, and led to false comparisons between (say) i5 laptops and i5 desktops.
With the A-series, there tends to be a new flagship iPhone range with a newly-numbered (or, at least, suffixed) A-series processor across the board every year, while the "old" models hang around as the cheaper options... and they're all ultimately ultra-low-power, battery-life-first processors anyway. I don't see that working with the "M-series" that will eventually cover everything from the iPad Pro to the Mac Pro...
I presume "M2" will use the A15 as it's base, just as "M1" used A14.
There is a chance of consumer confusion if a 2023 MacBook Air is on M3 and a 2023 MacBook Pro is on M2X, but to be honest I think Apple can handle that via marketing and product positioning and those who are performance-focused will already be looking at the CPU+GPU core counts and benchmarks so they will know going in the M2X is faster than the M3 on most/all tasks.
I mean the iPad Pro being on an A12X/A12Z with the iPhone on A13 and A14 didn't really confuse people about how much more powerful the iPad Pro was over the iPhone. Even the iPad Air on A15 was still lapped in a some performance areas by the iPad Pro on A12Z thanks to the additional CPU and GPU cores.
There is a chance of consumer confusion if a 2023 MacBook Air is on M3 and a 2023 MacBook Pro is on M2X, but to be honest I think Apple can handle that via marketing and product positioning and those who are performance-focused will already be looking at the CPU+GPU core counts and benchmarks so they will know going in the M2X is faster than the M3 on most/all tasks.
I mean the iPad Pro being on an A12X/A12Z with the iPhone on A13 and A14 didn't really confuse people about how much more powerful the iPad Pro was over the iPhone. Even the iPad Air on A15 was still lapped in a some performance areas by the iPad Pro on A12Z thanks to the additional CPU and GPU cores.
Well it's a "which came first: the chicken or the egg?" question since we assume both the M2 and A15 will use the same next-generation performance ("Avalanche") and efficiency cores. So if the 14/16 MBP launch next month at WWDC with M2...
This is very exciting. I'm super content with the M1 right now, but I'm thinking about what a Mac Pro will look like. That will be really something. May pick one up when that makes it appearance.
I just wish the iMac would ship sooner than what the rumors say, Fall.
19 days to WWDC maybe they surprise me.
oh and how about some 5K+ displays apple. please take my money.
19 days to WWDC maybe they surprise me.
oh and how about some 5K+ displays apple. please take my money.
What's that mean? I thought the way people were using M2 vs. M1x is the number referred to the generation, and the letter referred to the "series". Thus an M2 would be the next-gen low end chip, and the M1x would be the first-gen midrange chip. Given this, wouldn't the 10-core chip be an M1x and not an M2?
Not that I like that nomenclature There should be a different letter for each series, following by a number for the generation. Something like this would be much cleaner (gotta have an S in there somewhere, because all the carriage trade companies to do it):
iPad: A#
low-end Mac: M#
MBP: S#
high-end iMac: T#
Mac Pro: Z#
So this sounds like it would be an S1.
The definition according to cmaier is that an M2 would use an updated architecture, whereas an M1whatever would use the base architecture of the A14 or M1 but configured differently (more cores, larger LLC, LPDDR5 memory controller et cetera).
It makes at least as much sense as any other scheme, and focusses on CPU core architecture. These are only etiquettes after all, applied to hypothetical products. Given how many lines of comments that have been devoted to these names it would probably have been more efficient to simply write a description of the SoCs without any name attached at all.
I have the feeling that many people are getting too carried away with hypothetical chip names and completely loose the sight of what actually matters. As you aptly pointed out, what we should care about is whether the alleged prosumer 10-core (or whatever) chip is going to be based on the same architecture as A14 and M1, or is it going to be based on their successor. It would probably be helpful to have some sort of way to refer to the architecture itself somehow. We know internal names for the GPUs (the A14/M1 GPU is G13 for example), but I am not aware of any codenames for the CPU architecture.
Anyway, distilling it down to what really matters, both A14 an M1 are based on a Firestorm/Icestorm + G13 architecture, with M1 being little less than a "doubled" A14 chip, with some additional I/O thrown into the mix. He question is now whether the alleged 10-core chip will be based on the same architecture or not.
My completely uneducated guess is "no". I believe that the prosumer chip will be using a next-gen architecture and NOT Firestorm/Icestorm + G13. This is also how I understand @cmaier's post. I will briefly outline my reasoning below:
- Apple had ample time to build a bigger chip on the same architecture as M1, but they didn't
- Prosumer hardware would benefit from increased single-threaded performance, which Firestorm is unlikely to deliver in it's current iteration
- M1 is a really low-effort SoC, as it's a direct adaptation of the iPhone chip, just bigger and clocked slightly higher. It is a really incredible testament to Apple's progress in the area, because they are literally shipping fast laptop with an iPhone CPU, but I digress. The point is that M1 is a great first step, for for more serious applications, it would make more sense to have hardware that is better tailored for the needs of the desktop.
- Firestorm/Icestorm/G13 lack several features that one would expect in the prosumer space these days (vector extensions, hardware ray tracing, better ML performance etc.). Metal is already designed with hardware ray tracing in mind. Frankly, I would be shocked if we don't get hardware RT from Apple this year.
- The 8+2 core configuration is weird, as Icestorm cores are very small already and it doesn't sound like dropping two of them makes much sense, if any. This suggests that the efficiency cores in the alleged 10-core chip will see a significant rework.
To summarize, I think M1 is a "dead end" in the sense that it's an amazing chip and has delivered some great consumer hardware, but it's not suitable as a platform for building prosumer hardware. Furthermore, I believe that we will see more differentiation between mobile and desktop Apple Silicon in t he future. Right now, we have this:
A14 -> M1
And I think Apple will be moving to this:
Apple Silicon Arch Next ----> A15
|
----> Apple Silicon Desktop
That is, instead of using the mobile architecture directly, I believe that Apple will be deriving both the mobile and the desktop silicon from a shared architectural platform, but the mobile and desktop chips will both be better adapted for their respective roles.
Wait wait wait.
Are you seriously trying to tell me MR members are getting all worked up focussing on aspects that are completely unrelated to the functionality, because of some preconception they have about it?
SAY IT AIN'T SO.
Finger's off my M1Z Pro Tagmemics Fusion Advanced Plus!
Meh, I don't know if I care about that too much really, as long as they ship with SVE2.
Truly. If they're reworked so much that we only need 2 of them here, then why would the report claim we're still getting 4 on the M1 successor and even 8 on whatever will go in the Mac Pro? It's really bizarre.
Which is why I think (assuming the reports are true) that we will see more differentiated chips in the future.
According to Mark Gurman:
For the new MacBook Pros, Apple is planning two different chips, codenamed Jade C-Chop and Jade C-Die: both include eight high-performance cores and two energy-efficient cores for a total of 10, but will be offered in either 16 or 32 graphics core variations.
Are these the codenames for the SOCs or the CPUs?
I read this article recently, which shows that Apple is using the efficiency/performance cores differently based on the QoS of the process. I think that it's pure genius (because it actually prioritizes task, not only relative to other tasks, but also to save battery and avoid slowing the system). And it's further proof that Apple likely won't ship pro chips (whatever they call them) with less efficiency cores than the consumer version, as all processes with utility/background QoS run on efficiency cores only. If anything, I would expect them to have more, maybe even proportionally.
Just as a quick note, these things are actually explicitly mentioned in the official documentation: https://developer.apple.com/news/?id=vk3m204o