"Announce" and "replace" are not necessarily the same thing.
Apple has painted themselves into a corner with that "only one more to go, talk about that later" tease back in March. At some point Apple is going to have to talk about it later. Not necessarily ship but at least 'talk'.
Pretty good chance it is a "sneak peak" for the Mac Pro. We'll show you this now but it will be 4-8 months before you can actually order one. The Mac Pro 2019 would not be replaced by that, because the new one isn't shipping for a decently long time. But folks would know about the upcoming "orderly" transition between old and 'new' Mac Pro.
Note also though Apple gave themselves a huge 'out' with that "only one more to go" quip. That the Mini was already "transitioned" . So in the 'done' column. If they don't replace the last Intel Mini in 2022 , it doesn't matter because the Mini was already done. Lame? Yes. But it is a viable "dog ate my homework" excuse.
[ No , the Mac Studio isn't a viable replacement for Mini in dense computation rack contexts. It is 3x as big and not 3x as much computation (on CPU ) side. So it is a backslide too. What is missing a Mx Pro in the current classic Mini chassis. Period. ]
The M2 Mini isn't the problematical transition device. The M2 is still too much of backslide off the Intel Mini to be a viable replacement. The max RAM went up to 24GB . Which is still less than half of what the Intel Mini does. Half!! Anyone who has a workload RAM footprint of 32GB the M2 is still a *fail*. It is still a crippled backslide on video out also.
What is missing is a M2 Pro Mini. Probably still somewhat of a backslide from the old Intel 64GB, but not a comical less than half backslide. And the video out is not crippled. And the TB port count isn't kneecapped.
Yes no M2 Mini in 2022 would be lame excuse for product management, but that isn't an egg-on-the face transition blown moment. "Going to transition to Apple Silicon in about two years" promise absolutely squat about the pace of future Apple Silcon SoCs. Lots of folks presumed they were going to get iPhone A-series frequency updates, but Apple blew that out of the water last Fall. Only self delusion if still laboring under that now at this point.
If Apple announces in October that it is coming in "Spring 2023" and this is what it looks like and "let's watch a demo of it working", then that is probably good enough. Yeah technically they didn't finish in 2022 , but in the same two year period there was a world wide pandemic. That a smaller subset came in a Quarter or so late is not really the end of the world.
The much bigger hole that Apple would dig would be to say nothing. It would turn into a crisis like back in 2017 when lots of chatter was building that Apple was going to maybe hit 2,000 days of no Mac Pro updates. WTF are they doing? And then got the "we are going to do something" pow-wow meeting in April 2017. Same thing here. They would "have to" come out and do some "dog ate my homework" song and dance about how it was just going to take much longer than they thought. ( and if it is more than 6 months out that would be all you get. ). Most folks are going to interpret that "Later" comment in March 2022 to be "later in 2022". ( there was already a bunch of folks that whipped themselves into a frenzy that 'later' was WWDC 2022... which was not really well grounded. But again adds to their failed expectation management communication impact. )
Around April 2018 , Apple did a session where they said "not this year". That put folks off of trying to impact wild stuff detached from reality ( new MP at WWDC 2018 ). They would need to come back this Fall with something like "Much later ... as in sometime in 2023".
Personally I think they had queued up exactly what they had done in 2013 and 2019. Wanted to do sneak peak in June ( WWDC 2022). "stuff happened" and things slid into 2023. Not extremely far into 2023, but enough so that if they needed to stick to the "say nothing before a six month lead time" rule, that they had to move the "sneak peak" to Fall 2022. Decent chance October. If there is something seriously wrong then maybe Dec 2022 (and sliding much , much deeper into 2023). If very badly wrong then "recast it with a very broad 2023.
Very good chance the next Mac Pro is not going to make all current Mac Pro customers happy. Some are going to have a negative reaction whenever they do the announcement. If the new one is pretty far off the MP 2019 in slot count , 3rd party GPU support , ability to "raw iron" boot Widows , DIMM slots , etc then selling the current MP longer is, in the end, going to make those folks happier over the interim. Trying to rush a new Mac Pro to market when it isn't ready is only goes a bigger dust up with that customer base. For the narrower subset of folks they are targeting though it needs to be rock solid offering. There is already going to be a group of angry folks. Don't want to feed more folks to the mob that has going to gather around the product.
