Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Yeah this number is nonsense, the A15 is 15billion transistors and the A16 is 16billion, so it's not a much larger die and it's just a refined 5nm process so yields shouldn't be much of an issue.
Another reason might be to avoid a competitor filling order in that slot.
It's a huge competitive advantage to have access to the leading-edge node
 
I’ll bet that it was more expensive, but no where near 20% more expensive.
If it actually were that expensive, then other component costs must have come down enough to offset this. Apple is not going to suck up that much margin loss.
Yes they will in the short-term.
 
  • Like
Reactions: Sasparilla
Guess again. How cool would it be for Tim to stand on the stage and say their new iPhone 14 Pro Max has the same chip as last years 13 Pro Max? Apple basically needs a new chip every year to motivate consumers to buy the new model.
Tim Cook had me at Dynamic Island. That's all the motivation I needed. Everything else was just extra gravy :p
 
This news seems to be carefully managed Apple PR to prime our minds for price increases next year.
 
  • Like
Reactions: ksec
I thought so too, but the story so far seems to be that there was a big debate as to whether to have the phone run the processing, or have it all in-headset. I can't find the article now but I believe the rumors say the standalone headset side won.

We'll see. The only problem with having the phone run it, is how to transmit the data to the headset in real time.

It would make sense to do the processing in the headset. It's going to be highly annoying if the data can't be delivered accurately. I have Bluetooth headset that I use when doing yard work, and they drop out quite a bit at times. Anything VR that has that problem isn't going to sell well, duh.
 
A16 could be 24% more expensive than A15, but no way to be 2.4x (or 240%), the difference in process (N5P vs N4) and transistor count (or die area) are too similar to cost more than a double
"No way?" Do you have information about the yields on the new process? If the process is not as mature as planned, I can see lower yields (like only half the chips on the wafer meeting spec) driving up the unit costs. If less than half the chips are sellable but those sales have to cover the cost of processing the whole wafer it could lead to prices over 2x plan. People who really want the leading edge stuff seem willing to pay the premium. Low yields would also be consistent with the continued use of the previous chip for the non-pro models. Apple would not want to wait for improved yields so could go forward with the release schedule this way. (I have no visibility on this... but the assertion that there is NO WAY the new chip could actually cost more than 2x the old chip does not seem reasonable. This explanation is just one of the ways it could actually cost Apple more than 2x.)
 
  • Like
Reactions: hxlover904
Thanks to all the crazy people, who buy a new iPhone whenever their storage runs full of photos and videos.
Steve Jobs was right: You're the ones who drive innovation for all of us. Four nanometer for god's sake! 🤪
 
Why don’t they put the M2 into the iPhone Pro?
It wouldn't be efficient, power, cost etc. for what needs to be done in an iPhone. The M2 is truly made for the laptop / desktop and optimized for it, just like the A series chips are optimized just for iPhones and their (relatively) tiny batteries.

I think we now know another reason why they kept the A15 (albeit with the larger RAM) in the regular 14s. Makes sense. However it will be interesting to see how Apple does going forward...
I have a feeling you are right here. Either because of price increases (partly related to yield problems but partly cause TMSC is the only foundry with this tech working) and possibly not being able to produce enough A16's per wafer they restricted it to the 14 Pro's.

After reading that TMSC was raising their prices in general as the only foundry able to give companies this tech (i.e. have the market by the knickers), same for next years 3nm - I would not be surprised to see the A17 being expensive (whether there are yield issues or not) and that we have seen the iPhone SOC segmentation start which will never stop now. For the moment TMSC is in a bit of an old Intel power position over the market (we'll tell you what you'll pay) till someone else catches up. I'd expect price increases for the 15 Pro's next year as a result (of the TMSC tax), but we'll see.
 
Last edited:
Yeah this number is nonsense, the A15 is 15billion transistors and the A16 is 16billion, so it's not a much larger die and it's just a refined 5nm process so yields shouldn't be much of an issue.

And yet I got two downvote. Sometimes you want people to learn. But if history taught us anything, these people refuse to listen.

Edit: Surprised at least there are now quite a number of comments pointing out the obvious. May be Macrumors is improving. May be the world is improving. May be the truth will always come out in the end.
 
Last edited:
And people complain about them using last years chips in base model 14’s this year.

I hope this trend continues. Chips cost more, so Pro should get them because they cost more.

Bring on the hate. 👊😙👍
It’s like everyone forgot the first time separate iPhone product lines were introduced: the 2013 iPhone 5S and iPhone 5C

The 5C was an improved second generation iPhone 5 with the same processor, but new build material, slightly improved cameras, better battery, improved 4g antennas, and the introduction of colored iPhones.

Over time, we have become accustomed to the different iPhone products having more closely similar specs. What Apple did from a objective business makes sense.
 
The report is trash. TSMC itself slated N4 to be a refined version of N5 with reduced "mask counts, process steps, risks, and costs".

If the firm doing the cost breakdown was actually worth the pixels on our screen, they'd get more business examining why TSMC's N4 is twice the $/transistor as N5. That has implications for a multitude of chip designers worldwide. But the firm can't do that because it's not true.
 
  • Like
Reactions: ksec
It wouldn't be efficient, power, cost etc. for what needs to be done in an iPhone. The M2 is truly made for the laptop / desktop and optimized for it, just like the A series chips are optimized just for iPhones and their (relatively) tiny batteries.

That’s just naming, I guess.

They used the same chip in iPads and iPhones, now they use the same chip in iPads and Macs.

M2 has 2 more CPU cores and 5 more GPU cores, so they could just run them on a lower frequency and get about the same multicore performance.

And they could make the Pro iPhone actually a Pro and let it run macOS apps if connected to an external monitor.
 
(I have no visibility on this... but the assertion that there is NO WAY the new chip could actually cost more than 2x the old chip does not seem reasonable. This explanation is just one of the ways it could actually cost Apple more than 2x.)

That is not how N5P transitioning to N4 works. Ignoring current chip shortage and supply demand in balance, N4 should be "cheaper" than N5P. Hence the increase die size should in theory ( As Apple were doing cost projection ) cost the same as A15 on N5P.

Yield are also not calculated as such with Q&A from TSMC. Hence, yes. there are really NO WAY A16 cost 2.4x of A15 unless they did some clever financial engineering.
 
Yeah this number is nonsense, the A15 is 15billion transistors and the A16 is 16billion, so it's not a much larger die and it's just a refined 5nm process so yields shouldn't be much of an issue.

TSMC raised wafer prices by 20% and the additional EUV steps for N4 aren’t free. LPDDR5 costs double that of the older 4X.
 
It’s like everyone forgot the first time separate iPhone product lines were introduced: the 2013 iPhone 5S and iPhone 5C

The 5C was an improved second generation iPhone 5 with the same processor, but new build material, slightly improved cameras, better battery, improved 4g antennas, and the introduction of colored iPhones.

Over time, we have become accustomed to the different iPhone products having more closely similar specs. What Apple did from an objective business makes sense.
Thank you! 👊😃👍

I have nothing to add. Well said.
 
And people complain about them using last years chips in base model 14’s this year.

I hope this trend continues. Chips cost more, so Pro should get them because they cost more.

Bring on the hate.

Nobody is going to hate you for wanting to save some money on the standard models. It makes sense if the chips cost less and most people couldn’t care less or see the benefits anyway. The trend for buying last years Pro model is gaining momentum too apparently.
 
  • Like
Reactions: gund1234
If this article had come out before the iPhone 14 Pro event, I wonder how many people would have commented: "And Apple will pass that cost down to the user by increasing prices!"
 
  • Like
Reactions: gund1234
Nobody is going to hate you for wanting to save some money on the standard models. It makes sense if the chips cost less and most people couldn’t care less or see the benefits anyway. The trend for buying last years Pro model is gaining momentum too apparently.
I’m not worried about people hating me, I’m used to it. 👊😝👍

But there does seem to be a lot of hate and disgust on here towards buyers of the 14/14 Plus and towards Apple for making their decision to leave A15 chips in the standard 14 line.

I’m glad the buying of last years Pro models is gaining momentum. Especially for those wanting all the features they can get for cheaper. But a lot just want a brand new phone with a larger screen and lighter weight. Shoot, some just like the colors.
 
A16 improves so little compared to A15. Its basically a CPU overclocked A15 with no change to GPU. No way it costs 2.4x.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.