I think Apple Silicon has shown us that x86 doesn't have the monopoly potential it once had.
If you take that away, it becomes irrelevant (like many other ARM-based processors that struggle to be a good product because of compatibility).
Apple has a promising path for x86 liberation, but it is not there yet.
https://www.sciencedirect.com/science/article/pii/S138376212...
It's not the only ARM CPU with TSO support though, some server platforms also do it.
Also, can you name these other platforms?
Also, can you back up the information about most users not leveraging Rosetta?
https://share.transistor.fm/s/124f8bc4
> Also, can you name these other platforms?
Fujitsu A64FX, and https://en.wikipedia.org/wiki/Project_Denver
> Also, can you back up the information about most users not leveraging Rosetta?
Well I couldn't tell you that, but most people just use web browsers.
If most people just use web browsers, why so much effort has been put on Rosetta 2? Sounds very wasteful.
Also, A64FX seems to agree with my statement that an ARM with some x86 spice makes a better product than just an ARM.
And, if we're talking about making ARM mainstream, Raspberry Pi did more for the platform than Apple and deserves to get the good reputation.
It's mostly something that was needed in the first few years so you could run Chrome and Photoshop, but those have been ported now. It's mostly useful for running WINE but that's not super common outside games.
That said, a binary recompiler has a lot of uses once you have one: https://valgrind.org
What you're saying is exactly what I'm saying. Having that x86 translation layer helped them sell macs. It doesn't matter if it is not needed now (it is needed now, otherwise it would have been removed).
So, yes. Apple has a popular new band in town. But they still need to play some x86 covers to sell tickets.
As I said, they have a good plan, but they're _not there yet_.
When they'll be? When x86 translation is not shipped anymore. It's rather simple.