Are the power optimisations as good on Linux/ARM as on x86?
ARM chips use less power, that's kind of the whole point.
The processors do, that doesn't mean the desktop Linux distributions are well optimised for it. The available Linux phones have garbage battery life and a bunch of other issues.
This isn't necessarily as true as it once was. X86 has made a lot of ground in power efficiency and ARM has made a lot of ground in performance
Intel chips are still quite hot and use older process nodes which are less efficient. They have been pushing performance over efficiency recently as well. If this was AMD hardware on N5 I would agree with you, but sadly it isn't.
That's true in general, but Intel Atom is quite promising IIRC, and efficiency cores + improvements to their fabs should only continue to improve the situation.
I'm not saying the old logic of "ARM is efficient, x86 is fast" isn't still true, but it's becoming less true, and they're kind of converging to be similar chips but with different starting points (i.e., the needs are becoming more similar, and the differences are becoming lesser).
I'm not saying the old logic of "ARM is efficient, x86 is fast" isn't still true
Okay then I will say it. Apple Silicon is almost as fast per core than Intel and AMD. I am not talking just about x86 vs ARM in general because that's a fools errand. I am talking about Intel. That's also not an Atom chip, they don't make Atom anymore. Sure it is made of E cores but those are several generations removed from the Atom chips. It would actually make more sense imo if they used the 8 core version of that chip.
ARM chips use less power, that's kind of the whole point.
The processors do, that doesn't mean the desktop Linux distributions are well optimised for it. The available Linux phones have garbage battery life and a bunch of other issues.
This isn't necessarily as true as it once was. X86 has made a lot of ground in power efficiency and ARM has made a lot of ground in performance
Intel chips are still quite hot and use older process nodes which are less efficient. They have been pushing performance over efficiency recently as well. If this was AMD hardware on N5 I would agree with you, but sadly it isn't.
That's true in general, but Intel Atom is quite promising IIRC, and efficiency cores + improvements to their fabs should only continue to improve the situation.
I'm not saying the old logic of "ARM is efficient, x86 is fast" isn't still true, but it's becoming less true, and they're kind of converging to be similar chips but with different starting points (i.e., the needs are becoming more similar, and the differences are becoming lesser).
Okay then I will say it. Apple Silicon is almost as fast per core than Intel and AMD. I am not talking just about x86 vs ARM in general because that's a fools errand. I am talking about Intel. That's also not an Atom chip, they don't make Atom anymore. Sure it is made of E cores but those are several generations removed from the Atom chips. It would actually make more sense imo if they used the 8 core version of that chip.