Sorry, I meant to call it unified memory (not system).I'm not sure why you think M1 memory is "system" memory and Intel memory is not part of the system or vastly different in concept.
Sorry, I meant to call it unified memory (not system).I'm not sure why you think M1 memory is "system" memory and Intel memory is not part of the system or vastly different in concept.
Sorry, I meant to call it unified memory (not system).
I think for Apple it isn't just unified memory for the CPU & GPU but also for the other subsystems on the SoC. They made a claim during their Apple Silicon keynote last year that they designed the SoC so that all data for all the different subsystems used the same structures and all were available in unified memory including the Neural Engine and Image signal processors. I'm not sure how much difference that makes but they made a big deal about it.Intel SoCs have been using unified memory since at least 2012… Apple didn’t invent it. What Apple did however was to put a GPU-like memory controller on it and optimize the hell out of the cache hierarchy and the physical design.
I think for Apple it isn't just unified memory for the CPU & GPU but also for the other subsystems on the SoC. They made a claim during their Apple Silicon keynote last year that they designed the SoC so that all data for all the different subsystems used the same structures and all were available in unified memory including the Neural Engine and Image signal processors. I'm not sure how much difference that makes but they made a big deal about it.
I also expect that Apple can leverage their UMA better than other systems because they also write the OS. I wonder if Microsoft is looking into a new kind of driver design that would allow device drivers to better optimize UMA. Right now, device drivers are really single device designs and I would expect that trying to unify them under a single memory and cache structure is very difficult if not impossible.Oh, of course, I was just pointing out that the basic setup is the same as for any Intel or AMD SoC made in the last couple of years. All the processors on the SoC are connected to a common large cache which in turn is connected to the memory controller, so when a processor (CPU, GPU or whatever) makes a data request, it is served through that common system. This is what allows the processors to quickly and correctly see changes made by other processors (coherence). My understanding of these things is very basic, and there are probably important differences in individual implementations (I am sure that we have posters here that can offer more information), but I think it’s unfair to Intel and AMD engineers to trivialize their work. On the fundamental level, the unified memory approach as taken by Apple repeats the same principle as used by x86 chips for a and we should not get into the trap of ridiculing everything Intel does just because we are Apple fans.
This would deviate from Apples naming scheme from the A series. The reason why people assume the next chip is M1X is because Apple has consistently done an X variant of the A series line for years used in the iPad and iPad Pro. They also did a Z variant one year too which is why some think the third tier of chip would be M1Z.
I don't think we will be waiting two years for another Mac mini. I expect a new Mac mini later this year. It won't replaced the M1 mini though, I think it will replace the Intel Mac mini.MPB and iMac this year, Air refresh next year, then a Mac mini in two years from launch is my guess.
That, or a Mac mini pro, which definitely has a market I think.
This is very convincing as long as Apple will use only three tiers of Mac processors. In that case, M plus number plus nothing, or X, or Z, is perfectly fine.
My thought was that Apple could need four tiers. Then I lacked the fantasy what could come after „Z“.
Of course, my idea with the plusses only works if there are no more than four levels. You can utter „triple plus“ without issue, but four plusses would be bad.
I don't think we will be waiting two years for another Mac mini. I expect a new Mac mini later this year. It won't replaced the M1 mini though, I think it will replace the Intel Mac mini.
This is very convincing as long as Apple will use only three tiers of Mac processors. In that case, M plus number plus nothing, or X, or Z, is perfectly fine.
My thought was that Apple could need four tiers. Then I lacked the fantasy what could come after „Z“.
Of course, my idea with the plusses only works if there are no more than four levels. You can utter „triple plus“ without issue, but four plusses would be bad.
Agreed. I think whenever they announce the next chip (likely the M1X - or whatever they decide to call it), that will be when they also announce the higher end Mac mini that will come in space grey and have the same 4 TB ports as the Intel mini does now.I don't think we will be waiting two years for another Mac mini. I expect a new Mac mini later this year. It won't replaced the M1 mini though, I think it will replace the Intel Mac mini.
I'm only on my first MacBook, so I'm unfamiliar with how they rolled their lineups in the past. But it would seem to me they would want to differentiate the MBA and MBP a bit more. I didn't give much thought to the current MBP. Giving the 13" MBP a M1X this year would help. The MBA could catch up next year.
It’s a bit confusing because they have always sold two versions of the 13” model. One has only two ports, and is basically a MacBook Air, and the other has four ports and it’s more powerful. They updated the two port version last year with M1, but the “more powerful” 13” with four ports has not been updated yet.
It’s a bit confusing because they have always sold two versions of the 13” model. One has only two ports, and is basically a MacBook Air, and the other has four ports and it’s more powerful. They updated the two port version last year with M1, but the “more powerful” 13” with four ports has not been updated yet.
But it's kind of a weird spot in general for them right now. I guess I'm a bit surprised they didn't start with the new Apple chips at the top of the lineup and work their way down. Now it seems they have to introduce an upgraded version of the M1 to make the pro a pro.
No, just no...!"M1X" has become the common placeholder for the upcoming chips but I would like to suggest a different wording with plusses, i.e. M1 - M1+ - M1++ etc.
I could imagine that Apple will have four types:
- M1: MBA, Mac mini, as known
- M1+: MBP 14, smaller iMac, "Mac mini Pro": more cores, I/O, RAM
- M1++: MBP 16, bigger iMac, something like "Mac Cube pro mini": even more of everything. Maybe with upgradeability, with slots to add RAM and Apple-made cards with more cores of either GPU, CPU or neural?
- M1+++: Mac Pro (probably only with the M2-architecture around the end of 2022)
I would think that for me personally coming from an 09 Mac Pro that the SOC ssd would also accommodate virtual RAM in a way that would make it seem like 16GB was as good or better than the 24GB in my mac pro.That is basically the essence of M1's "magic" — it creates the illusion of needing less RAM because it's more agile in
It makes perfect sense to start at the bottom. Low-end chips are much simpler, so you are not risking too much (the M1 is not that much different from an iPad Pro chip), you don’t need to have all the intricacies (like multi-monitor support) worked out. By selling the entry-level model first, you are quickly increasing the adoption rate, give the software devs time to mature their ports, and gain valuable information while you finish up the feature more important to the pro user.
Well, I wouldn’t say “always”. The two-port 13” was introduced when, 2016? I am fairly certain that Intel’s stagnation and inability to deliver some SKUs was the primary reason behind the current fragmentation of the Mac laptop line. I’m sure that we will see some significant streamlining in the next two years (e.g. the two-port Pro being discontinued etc.)
From the latest rumours it seems we may be skipping straight to the M2X later in the year? I don’t see the point of releasing an M1X (based on firestorm & icestorm) even in July really when the 15th gen chip architecture will be just 2 months away with the September iPhones.
As for nomenclature I think M2X makes sense for a ‘base’ MacBook Pro chip, it could come in 6 and 8 (performance) core variants too, 35W TDP. I also think we could see ‘Lifuka’ incorporated into an M2Z, a higher end chip option, 8 cores or 10 cores or options for both with punchier graphics and maybe an upped 45W TDP to support it.
With the rumor that a 14" MBP is on the way, maybe they'll just discontinue the current 2 port model? That'd be kind of a weird ending to it, though.
But it's kind of a weird spot in general for them right now. I guess I'm a bit surprised they didn't start with the new Apple chips at the top of the lineup and work their way down. Now it seems they have to introduce an upgraded version of the M1 to make the pro a pro.
I can see apple only releasing a new mini every other year if it sells to well and people aren’t buying as many of the AS iMacs.I don't think we will be waiting two years for another Mac mini. I expect a new Mac mini later this year. It won't replaced the M1 mini though, I think it will replace the Intel Mac mini.
It also made perfect sense to use the existing chassises... to say "look what these chips can do" so it's easier to compare.It makes perfect sense to start at the bottom. Low-end chips are much simpler, so you are not risking too much (the M1 is not that much different from an iPad Pro chip), you don’t need to have all the intricacies (like multi-monitor support) worked out. By selling the entry-level model first, you are quickly increasing the adoption rate, give the software devs time to mature their ports, and gain valuable information while you finish up the feature more important to the pro user.