I don't know how to reply to this without just repeating myself. We seem to be talking at cross purposes. If you can't believe that Apple will invest in faster, higher wattage chips (despite what Apple themselves said at the keynote) then nothing I say will convince you.
Higher wattage than the A-series? I did not every say I believed they were restricted in that sense at all. That's you, not me.
Near as high a wattage as the current Xeon W-3200 and AMD Threadripper/Epyc. I didn't have to say that. Apple said it in their slide.
That reasonable large black area between where Desktops are now and where there is some "mac blue" shading.
Apple has no plans to be in that zone at all. They are self imposing a substantively smaller cap on themselves.
The part that "Macs with" is layered on top with is black ( no blue also). and the part with "Apple Silicon" is a lighest blue ( kind of sort of targeted area).
Apple Silicon is about pulling the desktops "left" far more than it is about group "up". About the opposite for the Notebooks ( far more about up than moving left ). Note also that the laptops are at edge of zone now and desktop are further out. Neither is exclusive direction but there is a core hub that they are targeting. That center of the hub is likely going to be a very real constraint on the Mac Pro SoC.
You are right we don't know exactly where Apple will take their new designs but this isn't Apple's first major transition. I seriously doubt Apple will go back to something like the 2013 Mac Pro—at least not initially.
The "at least not initially tagged on the end is indicative that even you can Apple's major goal here over time is to sell more Notebooks , Mini, and variations of the iMac. That will probably include iMac Pro class systems that have better horsepower than the current Mac Pro. And with less noise since they hav pulled back the upper end power limits.
They will very well be a very expensive high end Mac Pro at the edge of the space they are willing to target , but it is an edge product; not a core strategic product.
If you look at the previous two transitions,
The previous two transitions are largely misdirection when it comes to the Mac Pro space. The Motorola 68K was at a dead end. The product line was being shifted to embedded , not workstations. Every other major workstation vendor had either arleady started or had completed a shift off. Apple was tail end mover there and one of the last to "get off". They really didn't have another place to go. And Apple wasn't going to invest large sums to stay there either.
Ditto with Power PC. Apple didn't want to put money into staying on PowerPC. ( Sony and Microsoft got custom PPC chips out of IBM). Apple made several moves to neuter and kill off the PPC. First, was the instance on doing their own I/O chipset. Essentially, Apple had a dongle on the PPC implementations to do a Mac PPC. Second,
Lastly the other
huge difference is that Apple didn't have an OS waiting and mature on the other new target side. The Mac arm64 is going to get saddled with running iOS apps, but iOS is by a several orders of magnitude the dominate player on Apple arm64. PowerPC was pragmatically a "open, green field" given that only super expensive AIX was the operating system on Power before hand. It was trvial for mac OS to quickly blow past that in terms of active deployed units. x86 was different. Windows dominated there. That lowered the component costs as R&D spread over far more system desiner/builder/makers. It also lead to throwing in some compremises like emulated BIOS + EFI quirks and BootCamp support for Windows. Apple didn't control or make much money from Windows.
Both of moves were to a more open platform where there would be more robust OS competition and processors that were going off into niches. x86_64 isn't going off into a niche. Windows x86_64 is "abandon ware" by any stretch of the imagination. Neither is x86_64 Linux. AMD isn't dead (in fact the opposite) . Intel is far from dead ( although having problems haven't throw in the towel).
This move really isn't. Apple has already announced that they are locking down the boot option to just macOS. (maybe will get Windows but Apple and Microsoft are still hemming and hawing over that). Linux variants collectively have a larger than Windows deployment but iOS is a dominate player that Apple makes large sums off of . This isn't a "green field" move. It is more a "already are a huge iOS device farmer field".
Running iOS apps is going to a be significant priority this time. ( and iOS app all are going presume secure enclave , at this point a neural processor , and yes an iGPU wiith "apple metal" properties. ) . All the Mac SoC are extremely likely to be a superset of the iOS SoC baseline.
When Apple fully jumps off x86_64 will have lots of similarities to back when on the PPC track in classic dekstop user space if Windows on ARM doesn't take off ( which Microsoft probably can screw up). Apple jumping out of the x86_64 workstaiton business is a minor bump in the road for AMD/Intel because Apple is not a huge player here at all. Not even close ( in top 5 ). This is a bigger blow to Intel's laptop CPU package business than their workstation CPU package business.
Apple kept the basic chassis/design of their high-end professional system and just updated the internals to the new architecture (Performa 9600 to PPC 9600; G5 to Mac Pro). After all the drama leading up to the 2019 Mac Pro, and the inherent risks of losing professionals they've been actively courting back, do you really think they would risk upsetting that?
As long as Apple's workstation processors were on a share R&D investment plane with at least several other system builders then the costs could be spread out over a much larger number of units sold. ( even PPC had IBM workstations as an addtion. Wasn't huge but was something. Apple is going to have lots of nothing as alternative consumers of the CPU package for this relatively low volume product. That is not going to make this package cheaper. )
There is also the iMac Pro that will need a Xeon-caliber CPU.
Xeon spans a huge range from Xeond D , Xeon E (now W-1000 seiers) , E5 ( now W 2000 , 3000 series ) , SP etc. Not being able to "hit" Xeon is like not being able to hit the side of a large barn. That isn't a big hurdle to get over.
Same thing to lessor extent with AMD Threadripper -> Epyc . Both of those are going to be broader ranges than Apple covers. Can Apple get something that covers a subset? Yes. Are they going to cover all of the top end of the Workstation market CPU packages. Very problem no. Mainly what Apple will probably have is something that is just faster "enough" than what they have now in order to sell new Mac Pros. Pretty good chance that will be an even smaller subset than what is there now for workstation market.
So, yes, altogether I think there is a lot of incentive for them to develop tech that competes with Xeon systems like SVE2. And I think they've already got it on the drawing board if not in the fab.
Unless the laptops and iPad Pro need SVE2 I doubt Apple will bother. That is our real disconnect. You seem to think that Apple is going to go 100's millions out on a investment island just for the Mac Pro to stand "toe to toe' with with the upper end Workstation CPU packages. I don't. More likley Apple will extend the elements from the Mac line up to be 'good enough' for the bottom half of the workstation range.
Frankly Apple's investments in the Mac Pro have not been either consistent ( every year) or all that deep. ( major custom forks from rest of the Mac line up).
Neither are Apple's investments in the iOS , iPad OS space where the are far more systems with "hand me down' processors than "new ones each year. It won't be surprising to see the Mac roll out take two years because getting 3-4 Mac SoC all dribbled out one at a time. And then iterated on one at a time going forward ( e.g., 2-3 years to get to the next significantly upgraded one due to lots of 'single major update" tracking ).
Everything you're saying was said when Apple replaced the G5.
Again that is you saying that; not me at all. There was non one saying the PowerPC was going on to do gang buster sales in workstation space after Apple left at all. Not even in the slightest. In the G5 , Mac Pro 2006 era Apple did have any 3 year spans where made no investments in the Mac Pro class product at, let alone an almost 3-6 year gap (MP 2013 -> 2019 ) .
Intel (and AMD) had clear roadmaps for workstation processors after the transition to x86. That required zero, up front, direct, large investment by Apple at all. Not even remotely close to being the same. Apple is going off onto a very, very, very small island here in the Mac Pro context here. That hasn't happened in the past at all. In fact the G5 workstation was the exact opposite because was one place where the Power was
not retreating from. IBM was going to (and has) continue to make "big boy" CPU packages for Power.
They gave themselves two years for that transition and finished in (about) one.
Perhaps in some alternative universe.
"... The also dropped one of the biggest bombshells in Apple history: Apple plans a phased transition to an Intel-based processor architecture which it hopes to complete by 2007. ..."
MacCentral offered live coverage of Steve Jobs' keynote presentation to attendees of the Apple Worldwide Developers Conference (WWDC) 2005 in San Francisco, Calif., where he announced Mac OS X "Leopard" and a phased transition to an Intel-based computer architecture.
www.macworld.com
Apple basically gave themselves about a year (transition over by end of 2006). if count the remaining portion of 2005 it was 18 months. Not two years.
June 2006 they were pretty close to done.
"... Indeed, a little more than 365 days after Jobs’ bombshell, Apple has already moved most of its hardware to the new processors. All major Mac developers have committed to releasing versions of their programs that run natively on Intel chips—many native apps, large and small, are already available for Intel-based Macs. ..."
A year has passed since Steve Jobs dropped the bombshell that Apple would switch over to Intel-supplied processors for all its hardware. A little more than 365 days later, nearly all of Apple's hardware offerings run on Intel chips and major Mac developers have committed to making Intel-native...
www.macworld.com
The incrementally longer by a company with almost an order of magnitude more raw resources that could optinally be applied to this is in indicative that the Mac investments are definitely scoped to an area with real constraints. Prepped for two years and will take yet another two years to complete is substantially longer than the last time. ( and more A12X -> A12Z gimmicks of winking in something suppressed on the die to cover the longer cycle time. )
it maybe take a little longer than a year this time,
Because jumping to an incomplete line up. Day 0 both Intel and AMD had a full line. Apple may not have liked AMD's line up but it was up and running.
[quot
....but I'm confident they've already got the technology for a full replacement of their lineup. That isn't fanboyism that's good business sense and history.
[/QUOTE]
Fanboyism absolutely covers the premise that Apple's 10 year history over 2009-2019 screams huge , consistent investment. It does not.
Can Apple build something that is incrementally faster than the current W-3200 seires by 2022. Not really a big issue. Is Intel and AMD going to be sitting at W-3200 series performance in late 2022 - early 2023? Extremely likely not; at least not a broad set of metrics. It will certainly run native iOS apps faster, but embarrassingly parallel, large memory ( 500GB+ ) footprint workloads with a decent amount of pointer chasing. Probably not.
A big 'yes' to is this faster than a 2012 "Cheesegrater". But major sleepless nights for HP/Dell/Boxx high end workstation product managers ? Probably not.
