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.
iPhone 14 Pro and Redesigned MacBook Air Reportedly Stuck Using Technology Behind A15 Chip
- Thread starter MacRumors
- Start date
- Sort by reaction score
I do think the performance upgrade will be major, when I upgrade my trusty XS...
That being said, even my nearly 4y old phone is not that laggy. I only notice slow downs when aps are updating.
I might go for a 5th year and live with the freedom of not caring about my phone
That being said, even my nearly 4y old phone is not that laggy. I only notice slow downs when aps are updating.
I might go for a 5th year and live with the freedom of not caring about my phone
And customers will buy them anyway and they’ll make up a majority percentage of the Macs Apple sells in that year.
I’m on the same path with my iPhone X. I occasionally go to the Apple store to check out the latest iPhones, but they just don’t strike me as convincing enough to upgrade. Apple has essentially sold the same phone year after to year since 2017. But 2023 might be the year I take a serious interest in upgrading just for upgrade sake.
Last edited:
He also bought another $600 million in Apple the week before and currently holds $160 billion in Apple stock.
Clearly, he thinks the company is not long for the world.
People buy new devices every year, regardless how powerful or new they are.
And I’m sure folks are rofling over it. Doesn’t matter to Apple, though, they’ve made the sale.
Honestly that's all I care about. But also I thought Apple moved away from Intel because of their slow cycle ?
FWIW AMD would probably still have in house fab capabilities though if they hadnt had to spin it out for financial reasons because Intel used its market dominance to screw them over.
Most buyers don't.
But most buyers on these forums do.
And yes it's confusing to think that the chip in the 14 Pro will be behind the chip in the 13 Pro.
In the unlikely event that Apple is worried about a lawsuit over whether "M1x" is or isn't part of the "M1 family", then they can always call it the M2 (about which AFAIK nothing has been promised, including what process size it uses).
M1 or whatever is a product name not a specification, folks.
Possibly - or if they do make a "new" SoC for the Mac Pro it would be a good idea to call it something other than "M1/M2..." given that Mac Pros tend to have the longest product cycles of all...
I think another problem was that the M2 chip as previously rumored (significant single-core boost, extra GPU core) was shaping up to be a threat to the M1 Pro (certainly the binned 8-core version) so a spring/summer launched M2 Air would be in danger of outperforming the lower-end MacBook Pro on some workloads, which isn't even a year old yet (and, worse, is still suffering production delays).
I think people are right when they point out that the mass market outside such forums as this will be more than happy with a new design, colours and possibly screen for the Air, and an incremental performance boost from a souped-up M1 would be a bonus. Then maybe they can move the whole MacBook line to M2 together sometime in 2023.
From a marketing POV the "M1/M2..." marketing scheme that puts the "generation" upfront is a bit of a hostage to fortune - it only works with the A series because the primary A-series product is the iPhone, which gets an update every September, regular as clockwork. With a bunch of Macs on an irregular update cycle of 6 months to 3 years there's always going to be confusion over whether a "M5 Excessive" is better/worse than a "M4 Hyperbole" or a "M6 Incredible". Then, of course, they OSX-Lioned themselves by calling the second M1 chip out of the stable "pro"...
(So I guess the new intermediate chip will be "Mountain M1"...)
Not at the same size die level ( efficient die layout )
A9 and A10 were same process node, but the A10 got bigger.
A9
".... Apple A9 chips are fabricated by two companies: Samsung and TSMC. The Samsung version is called APL0898, which is manufactured on a 14 nm FinFET process and is 96 mm2 large, while the TSMC version is called APL1022, which is manufactured on a 16 nm FinFET process and is 104.5 mm2 large. ..."
Apple A9 - Wikipedia
A10
"... he A10 (internally, T8010) is built on TSMC's 16 nm FinFET process[1][9] and contains 3.28 billion transistors (including the GPU and caches) on a die size of 125 mm2.[10] .."
Apple A10 - Wikipedia
The A10 'bloated' up to "iPad Pro" A--X / M1 size and hasn't been that big since. [ The A10X had to jump to TSMC 10nm to stay in the 120-140mm^2 zone for a 'bigger A series' offering. ]
The A10 probably got some die space layout uplift because it was not dual sourced. but did Apple "effiecenly" save space to add substantive new features is on thin ice for evidence ( it is a bigger die). Did they make a bigger die that usually showed no increase in power levels. Yes. Did they save die space? No. ( got better at turning stuff not being used off. )
The problem with A16/M2 staying with TSMC N5 generation is that they have already done that. A14 (TSMC N5) and A15 ( pretty good chance was already on N5P. ). Apple took a die 'bloat' here also. A14 ( 88mm^2 ) and A15 ( 107mm^2 )
Apple A14 Die Annotation and Analysis – Terrifying Implications For The Industry
Despite TSMC’s claims of a 1.35x shrink on SRAM, Apple’s system cache has only shrunk 1.22x. This has far reaching implications for the industry. Designers cannot use increased LLC sizes as a crutc…
semianalysis.com
Apple A15 Die Shot and Annotation - IP Block Area Analysis
TechInsights released a die shot of the A15 and with the help of SkyJuice’s, we are analyzing it today.
semianalysis.substack.com
The notion that would keep N5 density and go to a > 107mm^2 for A16 and likewise grow the M2 past the normal "iPad Pro" die size is dubious when there is a N4 process with a 6% shrink readily available. ( MediaTek and Qualcomm are using it this year. Why would Apple skip it for A16 when had already bloated up on A15? That would mean more wafers to make the same amount of die. )
Apple hasn't stuck around for three iterations on the same process general node before. The notion that "well they have done two generation in the past"... yeah. But this wouldn't be two. They have already done two. The real illustrative past example would be looking for is 3. ( N5P gets no density improvements to "add more stuff". It is a power utlization update where can either incremental bump the clock speeds (and use same power) or use less power ( and keep same clock. ). Apple did a little of both of those with A15 )
GM206 (maxwell 2.0 512 shaders ) 228mm^2
GM107 (Maxwell 640 shaders ) 148mm^2
Maxwell 2.0 also largely went much, much bigger to get more power. Yes the new tweaks allowed the thermals not to spin out of control, but if use lots more die then should expect more performance if doing anything remotely competent. (more cache, execution units transistor budget allocation, more fixed function logic , etc. )
Why would marketing want to "save the M2" tag if the M1 is almost two year stale? Apple is about to hit they front edge of their "two year transition" timeline window sitting on the exact same "M1" they started off with. That is clearly not showing progress.
It isn't Apple who has a big bug up their but on the "M2 has to be some giant leap" notion. It is the tons of folks on these forums who have pushed the notion that Apple is insanely better designers than everyone else and M-series is going to take over the PC world . Apple is good, but they have always been riding on a 18 month Moore's law cycle to make major contributions to their year-over-year run up on performance. When that slide out closer to 24 (or 30 months) the pace will slow down. Apple isn't going to engineer those contributions back in and still keep the die costs in the same ballpark. (And yes Apple probably does have a die size and cost constraints on these dies. Pro/Max get bigger budgets but also likely have a 'cap' also on costs. )
Apple moved away from Intel for several reasons, the biggest being the fact that Intel’s power-hungry monsters barely worked in the designs Apple was trying to build.
Look at the 12 inch MacBook.
Super thin, fanless, small, horrendous performance that was out performed by iPads.
Or all the thermal throttling with the MacBook Pro and air.
And yes, i’m sure scheduling had something to do with it, since whenever Apple had a new design ready, they had to wait for Intel to release the processor for it.
Now Apple is on their own schedule, and they don’t have to wait for anyone to update the designs or whatever they want to do.
But this never meant that there was going to be yearly upgrades, even the iPad‘s processor never got that.
A9X: November 2015
A10X: June 2017.
A12X: November 2018
Not sure why anyone expected the Mac to be on a different schedule.
18-24 months.
You represent a very small minority of MBP buyers, though. Most people don't wait on specific node/fab upgrades-- or even perf bumps-- before upgrading. The ones that do mostly reside on forums like this
(Granted, I'm one of them. But we are in the minority.)
LOL, how else do you think performance increases without adding transistors? By Tim Cook waving his hand?
The data fully supports this claim. The entire semiconductor industry is based on this principle.
Apple went from 1 billion transistors in the A7 to 4 billion in A11. The A15 has 15 billion transistors.
Not only that, they have more than enough power for the average user. In many cases, even overkill. It’s nice that even the base model is so powerful (and efficient, at the same time).
Actually, you may be right. I was one of those people who didn’t care or wasn’t well informed when I bought my first MBP in 2010. I should’ve waited for the first i5 13” MacBook Pros in 2011. Or even better, the 2012 MBP with a better integrated graphics and longer software support. But I had to buy a laptop with an already outdated Core2Duo CPU in 2010. However, the performance was good during all its lifetime (more than 8 years).You represent a very small minority of MBP buyers, though. Most people don't wait on specific node/fab upgrades-- or even perf bumps-- before upgrading. The ones that do mostly reside on forums like this
(Granted, I'm one of them. But we are in the minority.)
Actually, I had to buy it because at the moment I had no laptop and I needed one, and I couldn’t wait another whole year. And for the people that have no laptop and need one, the upcoming 5nm M2 MacBook, or even the current M1 MacBook Air, are killing machines, and honestly if I were in that position, even knowing the 3nm are coming in the near future, I would buy whatever is coming this summer, regardless if it’s 5nm or the new chip ID is just a rebranded M1, I wouldn’t care.
Last edited:
Probably not The primary issue was not Intel market dominance. The major problem was that AMD Foundry services did not have enough customers for their fab. That was entirely evident when AMD got stuck with binding fab production contracts that made them make their stuff at Global Foundries (GF).
No viable customer base for the foundry is primary reason it is spun out. There was 'nobody' to share the increasing R&D costs with. Wasn't really when was spun out either, but the presumption was that new business would "flood in" ... and it didn't happen.
Extremely unlikely that new business would have flooded in if they had kept there would have been even less uptick at their foundry services.
Intel faces the exact same issue; it is just somewhat delayed for them. No honest effort in paying attention to contract fab business and they'll be toast at the top end of the scale in 5-9 years. Buying companies like Altrea and trying to "stuff" that increased workload into their own "inside only focused" fab doesn't really work well long term. A large enough breadth across product areas is probably going to require some fab process product lines that are good for others also. [ Wall Street hand-wavers were pushing Intel also to drop the fab business part and they were in the PC and server market dominate position. So dominate market position of those subset markets aren't the key driver here. ]
AMD's previous peak market share was 25% (around 2006) . When was AMD ever near 45-50% of the market? Yes Intel beat them down to a very low share in the 10 years past 2006. But that 25% wasn't viable long term. 60% wouldn't be viable long term either. And they never go anywhere near that.
The primary problem is that the cost of bleeding edge fab is growing too high for any single company to pay for. That why have to do some effort of "pooled sharing" for the funds to build them at scale. Even "contract" GF 'quit' at EUV/7nm.
Except it won't. Even if A16 (14 Pro) has the same cores as A15 (13 Pro), it will be made on a newer 5nm process (5NP vs. 5N) which offers either 20% better performance or 40% greater efficiency. And Apple can do a mix of both, so they can offer more performance and better battery life in the 14 Pro over the 13 Pro.
But the A16 will likely have new cores compared to A15 so the performance and battery life will be even better than what they would naturally get from moving to the 5NP process from 5N.
A15 should already be using N5P.
N5P offers -10% power at iso-performance or +5% performance at iso-power. We already saw those efficiency improvements with A15. N4 offers so little benefit, TSMC doesn't even disclose publicly the numbers.
Even if A16 uses N4, the benefits aren't appreciable. It's similar to when Intel slowed down with 14nm+ and ++.