That is absolutely
not the 'tick tock' model.
" ... Under this model, every new process technology was first used to manufacture a
die shrink of a proven
microarchitecture (tick), followed by a new microarchitecture on the now-proven process (tock). ..."
en.wikipedia.org
Completely skipping a complete microarchitecture generation is not a shift of focus from fab process 'wins' to microarchitectural 'wins'.
Intel applied 'Tick Tock' to their
entire consumer CPU processor line up for each release for several years. ( The server side moved slightly slower and would re-use the I/O support chips a bit longer. But again applied to the serve segment that reused the same baseline CPU core design with cache and feature augments. ). It is absolutely not a mechanism for dealing with low volume, 'niche' products.
Even when Intel was using "tick tock" though they did not throw 'tick' in the trash can (stop selling) when 'tock' came out. Throwing away chips on a yearly schedule very often doesn't particularly work well as need to recoup all the development costs in a relatively short period of time on a limited set of products. The "Ultra" and "Max" really don't have any Apple 'hand me down' product placements to use. The iPHone Pro Ann Pro die , now that a different size and makeup than regular one. Same problem, which is why stuff iPad Mini, and AppleTV and Neo are being pressed into 'hand me down' service.
Apple avoiding larger ( more expensive to design and construct) , lower volume chip packages has more to do with them not being profitable to make. Tick-Tock has relatively little to do with profitability and far more to do with control tech risk problems. Take a working system and change a limited scope significantly reduces the likelihood of running into more problems. That was a major contributor to how Intel went down the rabbit hole. The fab process rollout process got delayed and then tried to change a vastly higher number of things all at once. That proceeded to dig a bigger and bigger hole.
The Ultra itself was a compromise to control costs since the versions up until now have 'reused' a Max to get to an Ultra. [ Apple is being a bit opaque how the Max is constructed so hard to tell if M5 Ultra will be two Max subcomponent sets 'glued' together. ]
M1 Ultra , M2 Ultra , M3 Ultra. The M2 variant probably a bit a 'fuke' do in part to getting the line up with mix of pandemic , no "bigger than Ultra" solution (hence Mac Pro blows past 2 year transition timeline) , and N3 ( N3A --> N3B and fab roll out hiccups). The Ultra was always likely a odd-number or even-number sequence over the long term due to economics.
As for niche market.... Apple is having trouble keeping the Studio's in stock. Selling out isn't particularly niche.
( select a Ultra Studio at Apple Online store it and is saying Oct/Nov to pick up. ). In part there is no Mac Pro because Apple doesn't have any 'spare' Ultras to put into a Mac Pro. Yes, lack of memory is playing a role, but the major root cause of lack of memory is relatively high demand. "high demand" factors are not 'niche'.
There is decent chance Apple's resources are being spread thin also.
" ... At least a trio of companies are believed to be involved with the chip. Apple is said to be handling the overall design of the chip, while Broadcom is said to be providing some networking technology for it. TSMC is expected to begin mass production of the chip in 2026, using its third-generation 3nm process, known as N3P ..."
Apple Intelligence servers are currently powered by the M2 Ultra chip, and they are expected to start using M4 series chips next year. In an eventual move away from Mac chips for server use, The Information today reported that Apple is developing a new server chip that will offer even faster...
forums.macrumors.com
The R1 and possibly chips specifically for the smart glasses are also going to take resources away.
