Your estimates of announcement/launch date, model for first Apple Silicon Mac?

[richinaus]

Another factor in pricing is if they want their Mac market to grow or contract. Their Mac market has been very consistent in volume the last few years. (Which could also be described as "going nowhere".) They could have just kept going with a quite successful and profitable business. But they decided not to. And while Apple loves their profit margins, I doubt they changed their silicon strategy for the Mac with sales to less people as a goal. What would be the point of that? Particularly as services is a growing revenue stream, they want to add people to their eco-system, not grow more exclusive. And when it comes to sales volume, pricing is a huge factor.
Ergo, I personally believe the new systems will be priced to sell.

As Kotask pointed out, production costs are likely to go down generally speaking, but lets try to put some numbers on the table. Apple currently pays some $200 and up to intel for their chips, and $100 and up to AMD for their GPUs, rowing to $500 (and up) for the top end configurations.
Apple said they were making "a family of SoCs" for their Macs. Lets start the napkin math from the bottom:
A14x - Apple has typically made phone chips that are just below 100mm2 and iPad Pro chips that are just above. The A14 seems to continue this pattern, so lets assume that the A14x does too. Let’s put it at 11x11=121mm2. That would yeild almost 500 chips from a wafer. Lets ballpark Apples wafer cost to $10000 at 5nm, and yield with modern design-for-yield strategies to 80%. That would put their raw cost per die at 10000/(500x0.8)=$25/SoC. Obviously, this disregards all fixed costs but for a chip that is likely to be put in tens of millions of devices, it’s a starting point. Even if we double the cost per chip, it is still way lower than what Apple pay for intels low power Core chips. So I’d guess the low power draw MacBooks and possibly Mac mini will drop in price.
Next tier up (MacBook Pro + small iMac):
200mm2 SoC. Yields 300 SoCs/wafer nominally. Lets be a bit more pessimistic with yeilds and put it at 75%. 10000/(300x0.75)=$45 Lets round up to $50 in raw cost per chip. Now the volume of these chips is much lower. 5-10 million, depending on Apple pricing and market success making fixed costs a larger factor, and doubling to $100 is definitely prudent napkin math. However, that is still clearly below current chip prices from intel and AMD for the MacBook Pros.
Last tier, (large iMac, Mac Pro):
350mm2 SoC. This is the level that is supposed to improve on the PC offerings of 2021/2022. This would give a bit under 200 dies per wafer. Lets assume really bad yield in spite of binning and design-for-yield. A hundred usable SoCs per wafer would give a raw cost per die of $100. That’s still really low, but this is the chip where the fixed costs really start to bite hard. Assuming a volume of a couple of million SoCs, design cost+ mask set +... could still double the cost per SoC or more. But this is much less than what Apple pays for CPU+dGPU with dedicated memory!

The purpose of this exercise is to show that Apple can make even large SoCs for relatively low volume products, and still save substantially on their BOM. I hope, and actually believe, that they will leverage this to increase their market share.

Since when has Apple price their product to sell? They will just increase their margins if your figures on cost of manufacture is true.

 

[Yebubbleman]

I keep repeating myself in several threads now: Apple did not sell the same models of Intel and PPC side by side. Did they sell a 17” PowerBook G4 next to the 15” MacBook Pro Intel? Yes, they did. Did they sell the latter next to its 15” PowerBook G4 sibling? No, they did not. Discontinuation date of the 17” and 20” iMac G5? Jan 10th 2006. Release date of the 17” and 20” iMac Core Duo? Jan 10th 2006. And I could go on for all the other models: release date of the Intel version is on the day when the PPC version was discontinued. Please check for yourself on everymac.com.

I wonder where this notion comes from.

You are wrong. I know this because I was in the market for a new Mac in January 2006 when the Core Duo based iMacs and 15" MacBook Pros first burst on the scene. I remember laboring over whether to buy the G5 model or the Core Duo model and which side of the transition I wanted to be on (spoiler alert: I went with the Core Duo and not the G5, but it wasn't ). Similarly, the 15" MacBook Pro and the 15" PowerBook G4 did co-exist briefly in a similar fashion. The 15" PowerBook G4 disappeared first, followed by the iSight iMac G5s. But they absolutely existed and were sold side by side for the first couple of months. This notion comes from the fact that you don't have your facts and history straight.

You can cite everymac.com all you want, however, I can cite actual timestamped printout PDF files of shopping carts dated in January and February of 2006 also having been there (which it sounds like you were not).

It wasn't until the Intel Mac mini that Apple had started to discontinue the PowerPC models right as the Intel models were launched. They did leave sales open for the PowerMac G5 and iBook G4 in education and business specific outlets for a little while after the first Mac Pro arrived, but (a) that didn't last long and (b) that wasn't usual for them to do and was obviously designed to support specific markets that couldn't simply handle applications in Rosetta. They are not likely to repeat this with anything other than the 16" MacBook Pro and the Mac Pro (and even then it's not a guarantee that they'll do that at all this time around; though it really wouldn't surprise me).

 

[Jouls]

You can cite everymac.com all you want, however, I can cite actual timestamped printout PDF files of shopping carts dated in January and February of 2006 also having been there (which it sounds like you were not).

There is no need to get personal, Yebubbleman. As I never intended to get personal towards you. And no need to imply things. I lived through both prior silicon transitions. This will be my third. My recollection differs from yours, but your post got me thinking and I did some research on web.archive.org.

And yes, you are right and I was wrong. My apologies and thanks for correcting me.

 

[DHagan4755]

And yes, you are right and I was wrong.

Meh. In the grand scheme of things that doesn't really matter. The reality is there will be left over Intel Mac inventory that Apple will sell to people who still need the Intel architecture for whatever reason or for those who want to save some money & are low information consumers.

What matters is when this is all going to start to go down & what models will first see this new Apple Silicon architecture. That's what I'm most excited about. Right now the only rumors are Apple is announcing an Apple Silicon MacBook Pro 13" in November & the announcement date is pegged for November 17th.

 

[DHagan4755]

https://twitter.com/i/web/status/1293699651221192705

 

[EntropyQ3]

Since when has Apple price their product to sell? They will just increase their margins if your figures on cost of manufacture is true.

Possibly.
Then again, that seems like a pretty weak reason to go through this transition. Increasing margins on Macs doesn’t really make a whole lot of difference to Apples bottom line. From a business standpoint it makes a lot more sense to grow their platform. That doesn’t mean they’ll suddenly start producing low end crap - they won’t have to. They can keep their ethos and still be able to reach a wider market because they won’t have to feed their own and intels margins.
Obviously they will still build high end products, lowering the threshold for individuals and perhaps even administrations to be introduced to MacOS will simply help long term growth.

Another vector for growth is to provide compellingly superior products compared to what is available in Windows PC space. They are not mutually exclusive. But I can’t see the point of making this transition in order to remain a 7% player in personal computers effectively supplying the same product to the same people. I think their ambitions are higher.

 

[dmccloud]

In September & October events Apple made design refresh for iPadAir and the iPhone 12's, so maybe that hints to Mac redesign change in November. That is encouraging for 14" to 13" unibody redesign, which I hope it is. Will make no sense with the newly introduced devices to keep the Mac to the old boring unibody.

At the same time they updated the iPad Air design, they left the iPad design unchanged. I am not convinced that Apple would launch a new architecture and physical case design at the same time on the Mac, because that would leave far too many variables at play. If they utilize the existing chassis, they can focus on tweaks to the SoC, logic board, etc. since the case designs are already proven, especially with the new keyboards. Apple could also be saving the new case designs for 2021 and what models they will be announcing in that timeframe.

 

[Yebubbleman]

There is no need to get personal, Yebubbleman. As I never intended to get personal towards you. And no need to imply things. I lived through both prior silicon transitions. This will be my third. My recollection differs from yours, but your post got me thinking and I did some research on web.archive.org.

And yes, you are right and I was wrong. My apologies and thanks for correcting me.
View attachment 968397

I only got heated because you acted like I was imagining things.

Also, nice find. That definitely takes me back. I miss those days, man.

 

[Jorbanead]

Just speculation, but I doubt the ARM Mac Pro's are ready already as they require much more performance (graphics and otherwise). So I could see 2 iMacs (in diffrent sizes) and fingres crossed an ARM Mac Mini, as the 3 non laptop models.

I know Geekbench is not always indicative of real-world performance, but it still shows how powerful Apple's custom chips are with passive cooling. All results shown: Single Core # // Multi-Core #

Apple Silicon

iPhone 12: 1564 // 3703
iPad Air (4th Gen): 1582 // 4087
iPad Pro (Older A12z Processor): 1119 // 4617

Intel Macs

27" iMac (i9 2020): 1251 // 9035
21.5" iMac (i7 2019): 1116 // 6059
16" MacBook Pro (i9 2019): 1097 // 6877
13" MacBook Pro (i7 2020): 1240 // 4521

The A14 Bionic (and the A13 Bionic) beats out every single Mac in single-core performance. The A12z in the iPad Pro scores better in Multi-Core performance than the best 13" MacBook Pro you can buy today. Thats a 2-year-old processor in an iPad with no cooling. We can expect that the A14x Variant will at least score around a 7000+ in Multi-core performance if we base it off of previous X-variant gains. Assuming thats correct, that would be a better processor in an iPad than the Top-tier 21.5" iMac.

Imagine what they could do if they put that A14x chip in a desktop with an actual fan? They could ramp up the clock speeds, and get even better single-core and sustained multi-core performance. Apple is already there with their chip performance. They easily could release a MBP soon that would at least be at the same level as their current line-up but with all the other benefits of Apple Silicon. I would even argue that they could probably beat out current intel Macs by 20%.

If they can scale this type of performance up to xeon-class levels of performance until the insane thermal headroom the Mac Pro currently has, It's not out of the realm of possibility that they are close to a Mac Pro machine already. The BIG question for me is the GPU though. I'm very confident they can get CPU performance in that ballpark. Without a doubt, the Apple Silicon Mac Pro will be the very last thing they announce.

 

[SWAON]

Apple could also be saving the new case designs for 2021 and what models they will be announcing in that timeframe.

That's probably more realistic guess than expecting it now. We'll see

 

[OldCorpse]

I don't anticipate Apple cutting prices or banking extra money. First, because I don't believe AS represents much if any savings compared to Intel. Sure, they don't have to pay Intel prices, but on the other hand they have to spend a ton of capital - ongoing, not one time - on R&D to keep developing those chips - that's very expensive. And there is no easy way to recoup that expenditure, because AS will be a rather low volume run compared to anything Intel that has the entire Windows world to spread the cost of production on.

But even if we assume that Apple is saving money, I highly doubt they'll cut prices or bank the money - most likely they'll spend it to make a superior production with superior features, components and materials. That's how you grow your market, not by trying to compete on price with a different architecture. Low prices have never been the Apple way.

In short, I'd anticipate the same or even higher prices. How do we know? Look at what happened to the iPad Air. The new iPad Air 4 is $100 more than the Air 3 - given the iPad price that's a very significant price boost on a percentage basis from $499 to $599, a 20% price increase. And it's all with Apple Silicon chips.

Prices stay the same, or go up. YMMV.

 

[CWallace]

To be fair, the iPad 3's hardware upgrades were more than just the A-Series SoC and that likely impacted the Bill of Materials costs and therefore the MSRP.

That being said, I do agree with you that I do not expect significant (or even insignificant) discounting as I do believe Apple will add in new hardware features that will consume much or all of the SoC component savings.

But I do believe ASi will give Apple the flexibility to offer different product lines and pricing so we might see less-expensive models (sub-$1000) even if we do not see inexpensive models (sub-$500).

 

[Mikael H]

To be fair, the iPad 3's hardware upgrades were more than just the A-Series SoC and that likely impacted the Bill of Materials costs and therefore the MSRP.

That being said, I do agree with you that I do not expect significant (or even insignificant) discounting as I do believe Apple will add in new hardware features that will consume much or all of the SoC component savings.

But I do believe ASi will give Apple the flexibility to offer different product lines and pricing so we might see less-expensive models (sub-$1000) even if we do not see inexpensive models (sub-$500).

We don't necessarily even want "inexpensive": Nobody wants anything worse than disk performance comparable to modern NVMe drives, or colors and brightness comparable to an IPS screen if they can afford it at all. You pretty much need to pass the equivalent of $500 to get such specs, but not at all by much nowadays. An "iPad equivalent" baseline MacBook should be possible to achieve in the $500-$1000 span, with Airs and Pros as more expensive alternatives.