it is just a battery. Could be going into a MacBook Air, MacBook Pro 13/14, or Macbook Pro 16", nobody knows at this time. This is most likely not going into the first release AS Macs, as it is pretty late in the game to be filing for something that is possibly already being manufactured.
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.
Leaked battery suggests first AS Mac is Air or MB12
- Thread starter awesomedeluxe
- Start date
- Sort by reaction score
I just don't think it'll be called the A14 chip though. They clearly said they were making a family of SOC's for the Mac that will presumably be very similar to the A-series chips, but will also come with differences that are more specific to personal computing. Plus I don't see them putting out a 2 performance core chip for the Mac.
Exactly. I see the DTK Mac mini as the baseline for apple silicon Macs. That also uses the same chip with 4 performance cores. When apple last did their DTK machine, it was also a very underpowered chip compared to the chips they ended up putting out.
I see 4-5 groups of chips. Not sure how the big LITTLE will divide up, but I just assumed all will have 4 energy efficient cores.
8-core (4+4) All Laptops & All Desktops
10-core (4+6) Pro Laptops & All Desktops
12-core (4+8) Pro Laptops & All Desktops
20-core (4+16) 30" iMac and Pro Desktops
36-core (4+32) Pro Desktops
Last edited:
I see your logic here but I don't agree 100% with it all. The DTK isn't to see how well their software runs, its to make sure it runs at all and to debug it ready for release of the new Macs. That said, if it ran like garbage, Devs might be put off the platform.
Developers need more RAM than most, it helps when compiling I believe but while some Intel Macs still ship with 8GB, I think the AS Macs will need a higher base line for the Apple GPUs. Since they share system memory and from what I vaguely gather, need more of as their bus is wider than conventional GPUs, 16GB as a minimum makes considerable sense to me. So I agree with your conclusions, just have different reasoning.
This is absolutely spot on and well put. I'd wager the A1-6 variants never left the labs and that was always the plan. I also bet the same process will have been applied to whatever SoCs are headed for the Macs. The lower end range are ready for release, the Mac Pro equivalent SoC is a bit further away but Apple will have 5 years worth of unread M1-5 CPUs or maybe more they've been plugging away at in secret for some time now.
Actually if Apple only matches performance of AMD for example (we know they have an edge over most Intel offerings already), they will still be doing so at a significant cost per part saving over Intel CPUs. And the price of the AMD GPU. Yes that doesn't factor the R&D, but lots (if not all) of that is already justified by iOS device sales which make way more cash than Macs do. They probably have cash reserves that they can't afford to spend on anything other than R&D as it gets taxed when you move it around. Apple Silicon gives control over the update timelines as well as flexibility in design and improvement in vertical integration where Apple is already the industry leader by a big margin.
So an attractive, innovative model that matches an AMD equivalent in performance but costs less to build, can be priced much higher due its design (and likely much better battery life) will be adequate to appease investors and probably most tech journalists. I still think they are going to blow a few minds with what they release though, I just get a sense from them they know they are sitting on something pretty special this time.[/QUOTE]
Unless the A14-based Mac SOC with just 2 high performance cores is going to be just as fast (actually much faster in single-thread workloads) as the A12Z DTK with 4 high performance cores...
iPad Pro powered by 8-core A12Z (4 hpc / 4 hec):
iPhone 12 powered by 6-core A14 (2 hpc / 4 hec):
Not saying it is a sure a thing (I'd still bet on the base SOC for Macs to be a repackaged A14X), but if your prediction is built on the fact that the DTK represents a minimum "performance target" for the Apple Silicon Macs, then it would be perfectly achievable with an A14 SOC with just 2 high performance cores and 4 high efficiency cores.
Apple needs to completely blow Intel's CPUs out of the water. They also must be better than anything AMD has (as in the 4700G). Not blow it out of the water to the extent Intel's will be, but still better by enough to the inevitable "why don't you just use AMD off the shelf parts" questions go away.
As for the core count, there is no reason to make a dual HP core oart for the Mac. The incremental cost, as in the cost of the silicon itself, is trivial to go from a dual HP core to a 4 HP core. However, the 4 HP core can be used on the MBP 13"/14", where the dual core would be hard to justify for a Pro machine. There is also a non-insignificant cost associated with carrying extra parts (dual core HP, 4 core HP) that may seem insignificant, but in the manufacturing world, are very real. If power savings is paramount, take the 4 HP core part and underclock/undervolt it to get the power consumption values desired (the famous "binning"). You minimize R&D incremental costs, test/verification costs, and maximize volumes of the 4 HP core parts, as well as having the marketing advantages of advertising 4 HP cores even on the MB Air. Apple could create, for example a single 4 HP core part that runs at 2.0-2.2GHz on the MBA, and 2.7-3.0GHz on the MBP 13"/14". The MBA could even be undervolted, to get power way down, say below 10W, allowing for a passive cooling solution, and allowing for a smaller battery, or longer battery life. For the MBP, run it at full voltage, active cooling (i.e.. a fan) and let power consumption of the SoC go to 15W or even 20W.
Everybody is talking like the DTK is a development platform; its not. Note that it is called "Deveoper Transistion Kit". Apple never really intended for it to be a development platform, I don't believe, or at the very least, it would have had more RAM. They issued a 4HP core/16GB unit, because it is the baseline hardware that developers can expect to see.
The other point to keep in mind is that the AS SoCs do not, and may never, use hyperthreading (intel term) or simultaneous multithhreading (SMT), so the Apple SoCs will need "real" physical cores as a comparison to the 2 core hyperthreading Intel/AMD parts. So if the current MBA is a dual core (and the vast majority will be quad core i5s) hyperthreaded CPU, Apple pretty much needs a 4 HP Core SoC to match up, specs wise. Also, on top of that, Apple is now designing products that stay vaible (=useful) for longer, so a dual HP core in a laptop is almost unthinkable at the MBA's price point. they build one now, because they don't have a choice. When the AS SoCs arrive, they will.
As for the core count, there is no reason to make a dual HP core oart for the Mac. The incremental cost, as in the cost of the silicon itself, is trivial to go from a dual HP core to a 4 HP core. However, the 4 HP core can be used on the MBP 13"/14", where the dual core would be hard to justify for a Pro machine. There is also a non-insignificant cost associated with carrying extra parts (dual core HP, 4 core HP) that may seem insignificant, but in the manufacturing world, are very real. If power savings is paramount, take the 4 HP core part and underclock/undervolt it to get the power consumption values desired (the famous "binning"). You minimize R&D incremental costs, test/verification costs, and maximize volumes of the 4 HP core parts, as well as having the marketing advantages of advertising 4 HP cores even on the MB Air. Apple could create, for example a single 4 HP core part that runs at 2.0-2.2GHz on the MBA, and 2.7-3.0GHz on the MBP 13"/14". The MBA could even be undervolted, to get power way down, say below 10W, allowing for a passive cooling solution, and allowing for a smaller battery, or longer battery life. For the MBP, run it at full voltage, active cooling (i.e.. a fan) and let power consumption of the SoC go to 15W or even 20W.
Everybody is talking like the DTK is a development platform; its not. Note that it is called "Deveoper Transistion Kit". Apple never really intended for it to be a development platform, I don't believe, or at the very least, it would have had more RAM. They issued a 4HP core/16GB unit, because it is the baseline hardware that developers can expect to see.
The other point to keep in mind is that the AS SoCs do not, and may never, use hyperthreading (intel term) or simultaneous multithhreading (SMT), so the Apple SoCs will need "real" physical cores as a comparison to the 2 core hyperthreading Intel/AMD parts. So if the current MBA is a dual core (and the vast majority will be quad core i5s) hyperthreaded CPU, Apple pretty much needs a 4 HP Core SoC to match up, specs wise. Also, on top of that, Apple is now designing products that stay vaible (=useful) for longer, so a dual HP core in a laptop is almost unthinkable at the MBA's price point. they build one now, because they don't have a choice. When the AS SoCs arrive, they will.
Last edited:
1. What are you debating regarding the power envelope? You can already fit four perf cores in a 10W envelope (see the iPad Pro). I said this myself. The 9W APU in the Air right now is a four core part. We are literally looking at the battery and it's the same size and same Watt-hours... why would the power envelope be changed?
2. "since I am sure that the same SoC will be used on the 13/14 MBP" - why? Why are you certain of that? I laid out a good case for why the Air could use the iPhone SoC and/or the iPad Pro SoC, but you are just stating here that the Air will share an SoC with the Macbook 13/14 like it's as clear as the sky is blue.
It isn't. The MBP13 has a 28W APU. The Air's ranges from 5 - 9W, the iPhone's is ~6W, the iPad's is ~10W. The Air and the MBP13 using the same APU is a huge leap, and you can't just pull that conclusion out of nowhere and use it as concrete evidence that the Air must have a pinout to more than 16GB of memory.
Thank you. Can you point me at something that will help me understand how the mAH are different while the voltage and Watt-hours are the same - or was that a mistake in the article?
The most obvious change needed would be pinouts. I think their small, fanless notebook - whether it's called the Air or something else - is unique in that it doesn't necessarily need a pinout. Just as the developer mac mini just uses an A12Z with 16GB layered on, the MB Air could use an A14 with 16GB layered on.
So I'm not suggesting they would design a new dual core chip for Macs, I'm suggesting they would literally use the part that goes in the iPhone which it just so happens they will be manufacturing millions of starting in the Fall which is also when we think the Air is going to be released and wouldn't it be great if the Air just updated every time the iPhone did and no new R&D was required by Apple. I think Tim Apple might find that pretty neat.
The certification bodies lists the following information for a product with model A2389:
- 3C (China): charging voltage 13.05V and capacity 4380mAh
- UL Demko (Denmark): nominal voltage 11.39V, capacity 49.9Whr and 4380mAh
- KTR (Korea): provides a photo of the battery pack with no electrical specs
We clearly see 3 individual cells in the KTR photo. 11.39V / 3 = 3.8V. Max charging voltage for a 3.8V cell is 4.35V. And 4.35V x 3 = 13.05V. This is a known parameter for li-ion cells. Some people might confuse charging voltage with the nominal voltage of the battery pack.
We can be certain - this battery is for a future MacBook Air (13"). It has the exact same configuration, capacity, and space for the trackpad as the current battery pack.
awesomedeluxe:
1. Because the Mac SOC will NOT be the same. IPad/iPhone Pro SoCs do not have multiple USB/TB ports, as a minimum. They do have touch screen controllers/modules, which are not needed. They don’t have memory controllers for 16GB of RAM which will be needed. And Apple targets may not align with your logic. The biggest unknown right now is the actual performance targets that Apple wants to get to. Is Apple looking to beat the Intel powered equivalent Macs by 20%, 50%, or 90%? 20% could be done on with a 4 HP Core SoC, but does Apple want more performance than that? And lastly, Apple has stated that the Macs will be getting their own family of SoCs. Apple is going to develop a new Mac SoC for the reasons stated above. Customizing SoCs is part if the logic behind the move to AS in the first place.
2. With a view to section 1, above, and the close similarity between the current MBA 13” and MBP 13”, this isn’t a big stretch. In a bid to not have to keep designing a specific SoC fir every single model, Apple will design a few Mac SoCs, but not dozens. So to me, you design the “consumer” SoC for the MBA and 13” MBP. The MBA is the biggest volume Mac, so to maximize volume of SoCs, you design a common SoC. That means you make provision for 4 USB4/TB4 ports, and external RAM. You may not make use of them on the MBA, but they are needed for the MBP. On the MBA, maybe you only use 16GB of Soc RAM, and 2 USB4/TB4 ports. On the MBP13, all 4 USB4/TB4 ports, and 16-48GB RAM off the SoC. Good possibility that you use a higher clock speed as well on the MBP.
This is assuming that the SoC will only be a 4 HP Core part. To me, there remains the possibility that it could be a 6 or 8 HP core part.
1. Because the Mac SOC will NOT be the same. IPad/iPhone Pro SoCs do not have multiple USB/TB ports, as a minimum. They do have touch screen controllers/modules, which are not needed. They don’t have memory controllers for 16GB of RAM which will be needed. And Apple targets may not align with your logic. The biggest unknown right now is the actual performance targets that Apple wants to get to. Is Apple looking to beat the Intel powered equivalent Macs by 20%, 50%, or 90%? 20% could be done on with a 4 HP Core SoC, but does Apple want more performance than that? And lastly, Apple has stated that the Macs will be getting their own family of SoCs. Apple is going to develop a new Mac SoC for the reasons stated above. Customizing SoCs is part if the logic behind the move to AS in the first place.
2. With a view to section 1, above, and the close similarity between the current MBA 13” and MBP 13”, this isn’t a big stretch. In a bid to not have to keep designing a specific SoC fir every single model, Apple will design a few Mac SoCs, but not dozens. So to me, you design the “consumer” SoC for the MBA and 13” MBP. The MBA is the biggest volume Mac, so to maximize volume of SoCs, you design a common SoC. That means you make provision for 4 USB4/TB4 ports, and external RAM. You may not make use of them on the MBA, but they are needed for the MBP. On the MBA, maybe you only use 16GB of Soc RAM, and 2 USB4/TB4 ports. On the MBP13, all 4 USB4/TB4 ports, and 16-48GB RAM off the SoC. Good possibility that you use a higher clock speed as well on the MBP.
This is assuming that the SoC will only be a 4 HP Core part. To me, there remains the possibility that it could be a 6 or 8 HP core part.
I agree totally. I think a lot of people are thinking that Apple will be content to outperform Intel/AMD by 20%. I think they are looking for a lot more, as in a MBA that outperforms the current MBP 13” top end model by 50% or more.
I appreciate this analysis. Regarding the memory controllers and USB - I think these are not as big of an issue as you make them out to be. The developer Mac Mini has an A12Z inside but supports four USB ports and has 16GB of RAM. Were changes made to the A12Z? I'd bet not. It's not a stretch to me that the A14 could accommodate two USB ports and 16GB.
Your "common SoC" analysis in part two is probably correct. But you are foregoing the more obvious conclusion that the Macbook Pro 13 would share an APU with the MBP16, iMacs, and Mac Mini - probably the Bloomberg APU, with plenty of I/O. If the Air shared an SoC with the MBP13, what would be the point of the latter device? The Air is much closer to the iPad and iPhone in terms of both TDP and I/O needs.
But most importantly, you should remember what the Macbook Air is for. This is Apple's marketshare notebook, their opportunity to replicate the success of the iPhone SE with Macs. If they use the iPhone SoC they can make the Air really, really affordable. And it will still have excellent performance. And it can be sold with a four core iPad SoC for $200 more, which is pure profit.
You are talking about performance targets, but the Macbook Pro 13 is the performance machine. I am sure its performance will be exceptional. A good start would be not sharing an SoC with a 10W device.
You can get even an 6 HP core SoC to the acceptably low power level by running it at slower clock speeds, and lower voltage levels. Everybody is quoting 4-5W per core for each of the A13 HP cores. This is at ~2GHz. What happens if you take A14 generation cores and run them at 1.4-1.5 GHz? Perhaps run them at 1.1V instead of 1.25V? There are a lot of ways to slice up this pie.
Some changes were obviously made to the A12Z vs. the A12X, or else Apple would have called it an A12X. Were there changes to accomodate 16GB of RAM? Yes. Were 4 USB ports added? Unknown, but even if an external USB port controller IC were used, you still need logic to work with that controller chip, and to transfer data in and out of that controller’s USB ports.
in any event, there will be significany differences between the A12Z and what shows up in the first AS Macs. At the very least, we know that Apple will be creating a custom Mac SoC, and one based on 5nm instead of 7nm. I have no doubt that the external USB4/TB4 ports will be inside the SoC, and that there is a very good possibility thar there will be more GPU cores. Even with the new SoC for the MBA/MBP 13”, it will still be possible to drop the price in comparison to the Intel Macs.
i never said the MBP 13” will use a 10W SoC. On the MBP 13 SoC, you crank up the clock, say to 2.4-2.5!GHz, operate the SoC at 1.25-1.35 V, you coonect up the extra 2 USB4/TB4 ports, and you allow it to be configured up to 32-64 GB of RAM. Same SoC, just used differently. In both the case of the MBA and MBP, you save cost on the Intel CPU, Intel Chipset, T2 chip (will be inside the Soc), external RAM (the first 16GB RAM is on SoC, on the MBP, being Apple you charge through the nose for any extra RAM option), and a mass of glue logic (because most of it will be on the SoC). That also implies a simpler PCB, possibly with fewer layers. The cost savings will be there.
Some changes were obviously made to the A12Z vs. the A12X, or else Apple would have called it an A12X. Were there changes to accomodate 16GB of RAM? Yes. Were 4 USB ports added? Unknown, but even if an external USB port controller IC were used, you still need logic to work with that controller chip, and to transfer data in and out of that controller’s USB ports.
in any event, there will be significany differences between the A12Z and what shows up in the first AS Macs. At the very least, we know that Apple will be creating a custom Mac SoC, and one based on 5nm instead of 7nm. I have no doubt that the external USB4/TB4 ports will be inside the SoC, and that there is a very good possibility thar there will be more GPU cores. Even with the new SoC for the MBA/MBP 13”, it will still be possible to drop the price in comparison to the Intel Macs.
i never said the MBP 13” will use a 10W SoC. On the MBP 13 SoC, you crank up the clock, say to 2.4-2.5!GHz, operate the SoC at 1.25-1.35 V, you coonect up the extra 2 USB4/TB4 ports, and you allow it to be configured up to 32-64 GB of RAM. Same SoC, just used differently. In both the case of the MBA and MBP, you save cost on the Intel CPU, Intel Chipset, T2 chip (will be inside the Soc), external RAM (the first 16GB RAM is on SoC, on the MBP, being Apple you charge through the nose for any extra RAM option), and a mass of glue logic (because most of it will be on the SoC). That also implies a simpler PCB, possibly with fewer layers. The cost savings will be there.
You have efficiency cores running at 2GHz using a fraction of a watt. Why do you want six perf cores that aren't capable of running at 2GHz?
The A12Z debuted in the iPad Pro, not the Mac Mini dev kit.
I acknowledge that the MBP13 and Air could share an SoC.
The low end Macbook 13 has just two USB ports. That limited I/O means it can use an A14X. What makes sense here--still--is the MBA starting with an A14 and being upgradeable to a four-core A14X, and the MBP13 starting with an A14X and being upgradable to the 8 core Bloomberg APU.
EDIT: Just realized also that the MBP13 with two USB ports also has a max of 16GB RAM. So it can use the A14X with LPDDR5 layered on and not need a pinout.
Last edited:
They won't. iPad already has 4 performance cores and it's a tablet SoC on 7W TDP, putting just 2 in a laptop doesn't make any sense.
I doubt they go for the same thermal envelopes as they have now. Even tho we are talking about more cores, 28W TDP is what we have in the 13" and 45W TDP for the 16 is what it already has (even tho Intel bursts way higher in consumption).
Apple is aiming for longer battery life and in the chart they clearly stated they had desktop performance for way less consumption. Plus Apple Silicon will be on 5nm (far ahead of Intel's 14nm) so they don't even need to do something on the same TDP range as Intel to destroy them in performance.
My guesses are:
- 10W in MBA/MB: 50% more TDP than iPad + efficiency and performance improvements of new arch and new lithography (5nm) means something way more powerful than iPad. Also 50% more TDP is either 8 way more powerful cores, or just 50% more cores than iPad (12 Cores)
- 15W in 13" MBP: again, fits in the 12C range
- 25-30W in 16" MBP: 12-16C range
Last edited:
Sorry, got my iPad generations messed up.
You don't want to have 6 Perf cores incapbable of running at 2.5GHX/ You run them at lower speed on a MBA to reduce power consumption. You could also use a 4 HP core SoC, and run the cores at higher speed. Same difference.
You keep insisting that Apple will use the iPad Pro's SoC in the MBA, when Apple themselves have said that the Macs will use thier own family of SoCs. As much as what you are saying is entirely possible, I will side with Apple on this one, when they say there will be Mac SoCs, I take it to mean THERE WILL BE MAC SoCs. Not A14Xs. A14Zs, or anything from the iPad or iPhone lines. As I have written in other threads, Apple wants to really shake up the market. You don't do that by putting a tablet SoC into a full sized laptop. I am of the opinion that Apple is looking to beat the AMD 4700G by at least 20%, and the current top of the line Intel CPU in the MBP 13" by at least 50%, if not more, on the MBA. You can't do that with an A14X.
Where are you getting that the Macbook or Macbook Air has 50% more TDP than an iPad Pro?! Who says the iPad Pro has a 7W TDP?!??
Guys the TDPs for every processor that has ever been in the Macbook and Macbook Air are like two clicks away. Just do a little research please. You are making me sad.
And I hope you are right, because then we will get better Macs! Just don't be too disappointed if they end up selling the MBA with an A14... it will still be a good machine!
That's definitely for a MacBook Air and not a 12" MacBook. People need to divorce their crazy love for that machine with the facts in front of them.
Also, in the same way that the A12X/Z is still a member of the A12 family but designed for a high-end tablet rather than a smartphone, you're going to see, as Apple themselves have stated, a whole new family of SoCs that are Mac specific and not identical parts from iPhones or iPads. The fact that Apple used the A12Z in the DTK is no different than the 3.06GHz Pentium 4 that was in the PowerPC-to-Intel transition DTK.
Yes. They can and will do better than that.
Honestly, if they gave the Air that, two more Thunderbolt ports (which is concievable seeing as they're not going to be bound by Thunderbolt limitations built-in to the Y-series Intel Processors used in the current Air), give or take a TouchBar, they could absolutely do that and finally nix the need for a 13" Pro.
When most people say that a Mac will get an A14, they're not specifically mentioning the chip destined for the next iPhones. They're saying that something A14 based will be powering the Mac. Similarly, you could say that the SoC in the current iPad Pros and the Apple Silicon DTK is A12 based; it's not like people are saying that the SoC in the iPad mini 5th Gen, iPad Air 3rd Gen or iPhone Xr, Xs or Xs Max is what's in those devices. A14 will likely be broad in scope just like many of its predecessors (hell, the T2 chip in newer Intel Macs is A10 based).
Honestly, you make an A14 based SoC for the MacBook Air, and you likely can reuse the current cooling in that Air just fine with a 4+8 12-Core. Hell, you give it the heat pipe it currently sorely lacks and then you're really fine. All of the iPads (higher-end Pros included) have very little cooling the SoC. Hell, in most iPad minis there's nothing.
Apple will probably still continue to offer 8GB options to people. It's stupid, and they really ought to not, but they likely will for marketing/price point reasons. It wasn't like the switch to Intel really changed up the kinds of RAM capacities Apple was offering all that drastically from what the contemporary PowerPC models were running at the time.
I agree with this. They're going to have to return to the era of "Pentium crushing power!" that they were at before (a) to justify the switch away from the most popular consumer laptop/desktop computing architecture in the world that they fought and then later joined and (b) to prove to people their claim that it will lead to better products.
The shape of it is telling. It's definitely a 13" notebook. If it was a 14" notebook, it'd be larger. And if it was the 16" notebook, it'd be larger still. Apple designs the batteries around the chassis that they go in, not the other way around.
As for how soon the Mac it's going in will be launched? We're not seeing any Apple Silicon Macs until at least the launch of Big Sur. Given that the Public Beta only just launched (and historically that usually has a two and a half month lead time before release), we're not seeing our first Apple Silicon Mac until at least October, if not November. Rumors did peg the 13" MacBook Pro and 21.5" iMac replacing 24" iMac as being first with the Air being maybe early next year. Though that could change. I strongly suspect that Apple's Mac product lines will be given a shake-up not too dissimilar to the one that it had when Apple switched to Intel. There are way too many sub-15" Mac notebooks for how few there are of any other kind of Mac.
I think the MB Air or the MB survives, but not both.
I don't see an 8GB AS Mac, at least not in the first releases. Maybe that ends up fitting the MBP 13" being released first, and then the MBA/MB being released later and having 8GB. But I personally believe that the DTK defines the baseline machine specs. 4HP cores, 16GB RAM, USB4/TB4 ports. I would not be surprised to see the base SoC have as many as 8 HP cores, but won't be surprised to see only 4. Keep in mind, at this point, Apple has not shown any indication that it will be using SMT/hyperthreading.
A line up of MacBook & MacBook Pro makes sense, and is logical. A line up of MacBook >MacBook Air >MacBook Pro doesn't, at least to me.
I don't see an 8GB AS Mac, at least not in the first releases. Maybe that ends up fitting the MBP 13" being released first, and then the MBA/MB being released later and having 8GB. But I personally believe that the DTK defines the baseline machine specs. 4HP cores, 16GB RAM, USB4/TB4 ports. I would not be surprised to see the base SoC have as many as 8 HP cores, but won't be surprised to see only 4. Keep in mind, at this point, Apple has not shown any indication that it will be using SMT/hyperthreading.
A line up of MacBook & MacBook Pro makes sense, and is logical. A line up of MacBook >MacBook Air >MacBook Pro doesn't, at least to me.
To the best of my knowledge the A12X and the A12Z have the exact same chip design. As I understood it, the yield rates were too low at first and some GPU cores were faulty too often, so one of them was disabled in every A12X. One the yield was high enough and most SoCs came with all GPU cores fully functional, they enabled all GPU cores in the SoCs coming out of the factory and called them A12Z.
Apple engineers and VPs were pretty clear during the keynote and dev sessions that Apple Silicon Macs will have custom SoCs. No chance Apple ships a MacBook with a A14X processor, let alone the A14.
They also implied the new Macs would be more powerful than the DTK and the iPhone's A14 barely is (better single core, worse or similar multi core).
Besides no matter how powerful it proves to be on GeekBench, if their first Mac with the self-acclaimed super-powerful CPUs has a dual core processor in 2020 the press will not be good.
I don't think it really matters if it's an Air or just a "Macbook." I think they'll use the name Air because that's the model that sold better. What people are getting at is that they think the new ultraportable will be fanless, like the Macbook was. And I agree with that.
The needs of a 13.3" 16:10 fanless machine and a 12.9 4:3 fanless machine are very similar. It's hard for me to write off using the A14X in the Air especially now that the DTK shows any I/O or memory stumbling blocks are solvable.
I have heard this said several times now and would like to watch that part of the presentation again so I can try reading the tea leaves myself. Can you point me in the right direction?
When Apple said their release chip would be more powerful than the DTK, my only thought was "of course, because it will be an A14 design and that's an A12 design."
Unfortunately I don't have specific times (+ this is off the top of my memory), but the places I remember this being said were the main Keynote, the Platforms State of the Union presentation, and John Gruber's podcast (Federighi said it there).
Ehhh, with the way Apple talked during the Keynote I think they're more interested in "blowing Intel away" with perf/watt. I think the MBA chip to this end will be optimized for battery life while providing a 20-30% uplift in performance.
I just cannot imagine a world where the base MacBook Air has the same 16 GB of RAM as the base 16" MacBook Pro. If all Macs start with 16 GB of RAM, Apple will need to double the high end to 32 GB to differentiate them. See below, it may be time for exactly that.
History of RAM doublings on the base model 15":
Early 2006 -> Late 2006: 512 MB -> 1 GB
Late 2006 -> Mid 2007: 1 GB -> 2 GB
Late 2008 -> Early 2009 (first built-in battery): 2 GB to 4 GB
Late 2011 -> Mid 2012 (first retina): 4 GB to 8 GB
Late 2013 -> Mid 2014: 8 GB to 16 GB
As you can see, the 16 GB era has lasted substantially longer than any other in history. Granted demand for RAM is no longer increasing as fast as it used to, but it has to happen at some point - so why not Apple Silicon?
Apple makes a small fortune overcharging for RAM. They finally have everyone over a barrel with no cheaper alternatives. If they ship with 32GB basic they will lose a lot of upgrades. Remember they still have to ship an OS that functions well on 8GB for the next 5+ years.
Yeah I can't see any scenario where Apple starts the MBA with 16GB and MBP with 32GB. I feel like browsers are the only reason typical users would need more than 8GB for normal usage anyway, and they haven't dealt with that for years anyway. Also I wouldn't use the DTK RAM/storage as a baseline since the last transition the DTK had more than the MacBooks that came later.