Perhaps you should find yourself on a forum called PCRumors, rather than spending your time trolling on this one.
I'm very aware of how far back it goes, but Windows 10 is still supported on machines older than that, and Window 7 and XP run pretty much the same software. The OS is pretty much the least concerning thing -- it's the applications that are important....
It's not about 'nm' at this stage but transistor density, and TSMC are ahead on that, not as dramatically as the 'nm' nomenclature would suggest, but still ahead.
The point is that corporate users care about compatibility primarily. Apple isn't great on that (which is why enterprise has never been a strong segment for the Mac) but Windows is. Apple has other advantages- like hardware/software integrations, lack of bloat, and committed developers. If Windows abandoned backwards compatibility, there is almost no reason to use it- and it loses one of its biggest advantages over Mac (the other is cost and diversity of hardware, but that applies to Chromebooks too). Meanwhile, Apple would still have its advantages. And writing for a new architecture does have ROI- it has allowed Apple to move to a new, better platform 3(!) times.
Transistor density is very variable depending on which part of the SoC you're measuring it, therefore it can't be an accurate test of how good the fab process is either. An ALU wouldn't have the same transistor density as a cache, for instance.
Same reason we're not still travelling by horse and steam ships, probably.
The original Motorola 68k is closer in time to vacuum tubes than it is to modern CPUs.
Every report on fabrication I've ever read disagrees. Even Intel themselves say they have hit process parity and aren't planning on being able to match and then truly overtake TSMC (provided both hit their roadmaps and neither stumble) until 2025 and 2030 respectively. That's when they say they'll get process leadership back - after 2025.
I disagree big time on ROI, it hasn't allowed it to move forward at all in corporate America. I'm the guy where I work that has to inform upper management on ROI, and I agree with most of corporate America.
By that Logic why use anything then? It’s all gonna go bad eventually. Of course Apple silicon, as we know it, won’t last forever? Same with x86. But surely apple silicon is going to be around for the foreseeable future.
And faster system bus, which was a bottleneck as G4s got faster. I have an (upgraded) G4 Cube with a 1.8GHz CPU on a 133MHz system bus. Which is a shame.
