....
How many processors do you you think Apple needs to produce? Consider, for a family of current generation iPhones and iPads, they only use two processors at slightly different clock rates.
The use of "hand me down" and minor clock speed tweaks to cover lots of products is the point. Apple only has 2 maybe 3 active lines to cover the iOS-tvOS-iPadOS-Speaker space. That's is about 10-12 products with just a fraction of the new instances each year. Large majority here is "older stuff" into newer products.
The current Mac line is far more spread out.
At least four , probably five if not six. It is far more than just about tech CPU benchmark scores and up/down clocking. Each one of these is different die which would need its own validation and certification. It isn’t like it is 5 “moonshot” efforts, but it far more work that Apple has done for far lower unit volumes produced. The “un-cores” stuff matters.
1. Mac Pro / iMac Pro — high I/O bandwidth , lane count ( multiple discrete GPUs , etc. ) and core count. ( very low need for > 2 low power cores. If no macOS scheduler changes then count capped around 32. May need SMT/Hyperthreading to be competitive ) Possible sacriice iGPU for more those previous two items ( or at very least minimize its footprint ). quad digit ECC GB RAM capacity handling ( and appropriate MMU caches and structures for that).
2. iMac 27” — mid range I/O bandwidth, lane count ( one discrete GPU and 4-5 20Gb/s ports ) and core count . low to mid range iGPU. Triple digit GB RAM capacity handling.
( possible chop down core count for iMac. 21-24” models if have mid range iGPU )
3. Mini , MacBook Pro 16 — bigger still iGPU (than above) , mid-range I/O , lane count ( one discrete GPU although Mini probably doesn’t use that. ) . Double Digit RAM capacity handling.
( possible iMac 21-24 model with no dGPU options.
dragging the Mini into the mobile targeted processor space will probably open a gap versus what Intel / AMD options for 10-20W more. )
4. MacBook Pro 13” , MacBook Air. biggest iGPU ( no discrete GPU lane allocation) , double digit RAM capacity handling
5. Macbook ( über , ultra light . always connected ) small tweak on iPad Pro : maximize SoC . Modem integrated ( or interface is Apple still doing discrete ). Just enough for ports I/O ( no possible dGPU ) [ This is Apple’s Captain Ahab chase for thinnest, lightest laptop. ]
Something like “chiplets” isn’t likely going to solve the issues items 4 and 5 bring up. It isn’t doing it for AMD now and likely won’t for Apple later.
Apple could possibly toss the Mac Pro / iMac Pro category away and drop one that way. could merge what is left of the Mini and iMac into one. And the MBP 16 , 13 , and MBA into one. that would get down to three. But is the Mini-iMac really going to be completive across the whole range with what AMD (and Intel ) have?
So that would be a around at least doubling of the number of active designs they'd need to be working on. Yes, there would be overlap in the basic core implementations ( at least for the non Mac Pro / iMac Pro if try to be SMT competitive). But there is variations on I/O that will some individual work that would need to be done.
If Apple only does one laptop and one desktop design and plays with just clock speed to stretch that over 7-9 products then they will probably loose. ( That very probably won't meet the need). Even 3-4 over 7 products is quite thin given what the competitors are doing.
On the other hand if doing "custom" processors to get them out being "painted into a corner" could end up with six if try to get the Mac Mini out of its corner ( and don't "double duty" it with an iMac or MBP 16" solution. )
Apple cold 'punt' the desktop/workstation versions to a 3rd party to do ( Ampere or some new upstart ) but that wouldn't be "control for control" sake. That also may not get them the pricing ( volume of those sold to other players in same general market ) or execution of new product that was significantly better than both Intel
and AMD .
I don’t think so. Apple’s made it clear that they see their future in iOS type products. If anything, they would grow iOS, but, in that case, they have an iPad that’s cheaper than any Mac.
Apple put effort into the iPhone SE 2020 version because iPhone penetration has stalled in developing country markets. Apple is running into a growth problem because they have priced their stuff out being affordable to the vast majority of people on the planet. Not going to get product growth if folks can't/won't buy the product. Similar with iPads where Chromebooks were eating the iPad's lunch until Apple come out with the "often on sale under $300" iPad model. ( iPads aren't always winners down market. It has had (and continues to have problem in some areas).
Similar issue is percolating with Macs. It is going to get even worse if AMD and Intel engage in a protracted price war on mainstream desktop CPUs. Being able to build "thinnest and lightest" laptop with new Apple processor isn't going to mean much there.
If Apple is going to toss aside Intel
and AMD then they have to out compete both of them. Outside of the lightest of light laptops they have tons of unproven work to do. Once plugged into a wall for power... Apple is bring what to the table?
If "iOS is all that matters" is the future then they should just do an 'iBook' or iPadBook and put iPad OS inside clamshell. ( basically an always attached keyboard versus the $300 external one.). Likewise Apple could retreat the Mac back into just laptops and maybe an iMac 24" or Mini. But would the Mac ecosystem survive on just that?
