Saved money overall? How?
- By cutting out the middle man of Intel / AMD and commissioning TSMC directly. ASi can leverage much of the R&D that would have been put into A-series development regardless (CPU / GPU core designs etc.).
- By increasingly being able to merge iOS and macOS.
There are two active versions of macOS.
With lots in common. And the x86 version is increasingly going to be in a holding pattern, missing new ASi features.
They had to invent Rosetta 2 .
How many billion did that cost?
Apple is running around trying to help some app development stacks to re-optimize their GPU libraries around the new Apple GPU semantics and specifics. The costs might go down in several years when Apple permanently stops all macOS on Intel development and finishes sponsoring/enabling the transition over to extremely highly optimized Apple GPU stacks, but that is likely several years out.
I've got no idea about any of that, but surely Metal makes much of it transparent to developers? How much sponsoring does it require?
Apple is about to ramp out yet another whole operating system on rOS/realityOS layered on the shared stack. The AR/VR products should be a new revenue source but still have an underlying OS on top of a functional skill matrix that is stretched over multiple OS objectives. The organizational overhead for keep the function matrix is likely just about as high. There has likely been no huge drop in headcount.
I was referring to Apple manufacturing their own SoCs directly, rather than paying for Intel / AMD's profit margins. I never said Apple don't spend money in other areas.
If this is suppose to be 'saved money' due to absorbing Intel's profit margins into Apple's, then that is likely not as high as most fans are spinning it to be. Apple has smaller scale of sales and have to stick to leading edge process nodes to stay competitive. Neither of those two is going to increase net profits.
This is an interesting point, and why I added the 'presumably' qualifier when I said they saved money with ASi. It would be interesting to know what Apple pay for e.g. an M2 Max chip. Though even it were $1, they would still charge as much as they possibly can for their computers (as any business would and should), so the fact ASi Macs aren't cheaper than their Intel equivalents doesn't tell us anything.
Apple additionally threw over $1B at insourcing the cellular modem that isn't paying huge returns right now either. Qualcomm isn't stumbling on making progress on better 5G modems so Apple isn't going to trivially blow past them anytime soon. The Silicon business as an aggregate whole has some substantive "return on investment" problems to the point it is not just a pure cash cow. The "mountain of cash" notion coming from that subdivision is likely overblown.
Long term a couple years after Apple completely dumps Qualcomm and Intel there might be an upswing in 'savings'. That hasn't happen yet though.
As above - I understand Apple have costs in other areas.
Apple doesn't need to swallow the loss. It isn't a requirement. It is a 'wish' that some are trying to project onto Apple. However, it is not an actual requirement.
It
is a requirement, unless Apple want to go the way of SGI. Sure, people will pay a certain premium for a Mac Pro over a workstation PC, but the market isn't completely elastic. At a certain point, even
@maikerukun gives up and buys a PC.
The workstation is very expensive. Less people buy it because it is so expensive. So charge more money to fewer people to keep it around . Rinse and repeat. That is basically a prototypical pricing death spiral. It doesn't work well long term.
There is a pretty good chance that Apple raised up the Mac Pro entry price floor to a level they think their base SoC will get covered for costs. Yes that shrank the number folks they sell Mac Pros to , but it is also an opportunity get off the spiral , by putting more value into the system to stop the bleed in base size. I think what several folks are overestimating is how much difference change in SoC that $6K floor actually is going to buy. It is not a "whole the bulk of the SoC out and start over" coverage.
Surely what you described literally
is the spiral: raising the cost = less buyers = need to raise costs? The competition is also putting 'more value' into their systems each year - it's called progress. ASi desktop Macs are power efficient, but are certainly no more performant than the competition (other than in areas that leverage the Media Engines).
And if substantively change the semantics of the SoC then have pragmatically re-forked the OS and app layer also. So no long term 'savings' there either. And that is no saving for both Apple AND the macOS developers. ( not all macOS developers have Apple like margins and "mountains of cash"). Impacts on the ecosystem as a whole is a relevant factor. Myopically pointing at Apple's money pile size is pragmatically mainly just misdirection.
Getting macOS and iOS on the same page has got to lead to cost savings, surely?
Yes, some of the hypermodular folks are going to 'run away' , but all Apple has to do in enough replacements to fill the gaps. Frankly, Apple already ran off a decent fraction of the hypermodular folks, because they are also in the "not going to pay $6K" group also. It is the size of the non-intersection that actually matters.
Fair. I don't get the impression Apple has cared about that market in a long time.
Yeah, because that is pretty much the way can build up a mountain of cash. If have several products that are busy throwing cash away ... you don't accumulate much.
Not only does Tim Cook see it that way. Major stockholders like Warren Buffet see it that way also. jobs wasn't trying to give away money either ( The norm was that Apple didn't hand out dividends under Jobs. ) . "Tim Cook" this and "Tim Cook" that is largely just misdirection also. This isn't coming from just one person and not a recent leadership change either.
In fact, most other Apple customers will likely see it that way too. Why should iPhone/MBA buyers subsidize Mac Pro buyers product purchases. It amounts on a tax on the much larger set of customers on the far lower end of product price scale subsidizing the the upper 1% so that the upper fringe can keep more money in their pockets. Doesn't sound like something that is going to over popular with the masses.
Agree with most of this. A Mac Pro that is being subsidised indefinitely is not something to wish for. If there's no business case, it will be permanently on the verge of being discontinued, and hard for Apple to justify putting time / effort / money into its development.
Disagree about the cross-subsidisation by iPhones etc. Ultimately, the price of an iPhone is what it is. Customers have no awareness of how the price breaks down in terms of costs, and how Apple use their profits to fund different areas of their organisation. If Apple felt that an ASi tower workstation was 100% essential to the macOS platform, and to be able to sell one at a price the market would bear they'd need to eat some of the cost, then that's what they would do. The costs would be like macOS itself - an indeterminate proportion of the purchase price of a new Mac.
