Two things…
First, you don't need to be Kreskin to grok that Apple is working on the successor to the M1 MBP Pros. They've been doing that since about, oh, 4 months before the M1 shipped. Or more. It kinda amazes me how folks don't seem to understand how computer engineering works. Apple makes these things… entirely… they have the ability, in house, to fab prototype boards AND prototype wafers… they've been running M2 since before the M1 shipped. (At least some of their folks have been, "secrecy" and all.) And, there really shouldn't be a need to mod the chassis yet. Even with M3/3nm, I don't see any great change to be made in chassis, there just isn't much necessary left.
Second, I think—"IMHO"—it is fairly important for Apple to keep a cadence in releasing these M- chips. They need to show that they're keeping out ahead of Intel/AMD, or the entire thing becomes a farce. Yes, the M1 was a HUGE leap forward in performance… but it also shipped 18+ months/a year ago (for the Pro/Max)… Intel/AMD have not been standing still. Apple cannot fall behind… they can't even stand still! If they fall behind Intel/AMD (like they did after the PowerPC switch), it won't be good. And the viewpoint that "well, speed is less important than power consumption/heat" I don't think entirely holds water, and is pretty much directly contradicted by Apple's PR messaging (which is that Apple is providing BOTH). If Apple sits on releasing the M2 Pro/Max, and doesn't show that they've gotten their legs under them—are able to revise these chips quickly, on a similar cycle as Intel—it would effectively look like Apple is admitting that the "complaint" against Intel not being able to deliver in a timely manner, as needed by Apple for their product line ambition, was somewhat wrong… because Apple isn't delivering on a year-over-year cycle either.
I've stated it before, but I think Apple is working internally to get the "cart" back behind the "horse"; the M1 was effectively a muscled up A- series chip, that then got subsequently bulked up for the M1 Pro/Max/Ultra. That's not how you want to design your chips… you want to design your "leaps" first, and then back-port. That's the way to make money, because you're always skating to where the puck is going… rather than just expecting (read: praying) you're going to end up where the puck ends up. (And that's exactly what Intel got lazy and did… and it cost them.) I think Apple is really working on the M3 Max/Pro/Ultra core and fab'ing right now… that's where they're focused internally. The M2 Pro/Max/Ultra is still a 5nm afterthought. And it is possible that they are doing the business calculation of "can we just stretch the M1 Pro/Max out until 3nm is ready and then jump directly to that??", but I hope not. Again, IMHO, cadence is important here: showing everyone you're on a cycle, that you can meet your schedule, and that consumers can rightly expect performance improvement on a rational timeframe. 18 months to 24 months per bump is just a bit too long, we've been so Pavlov-trained for a 1-year cycle… and Apple holds much of the blame for that!
