>>4544
of course, but efficiency is also a metric and especially important for portable devices
>most tools are x86
Good thing you brought this up. I don't really understand, if they port windows to it, and the programs you run on Windows that aren't supposed to do any sort of low level dealing (e.g. dealing with peripherals, ACPI queries that sort of things) and are purely in the loop of file access -> process file -> save file, like video editors for example, who only talk to the kernel - why would this be an issue with those? Shouldn't the kernel if they ported it properly, behave in nearly the same way?
Apart from this if in the future for some reason arm becomes mainstream everyone is going to recompile their shit for that architecture.
>There are physical limits to it
For laptops yes i suppose
For desktops and servers the only limit I can think of is the area of contact of the heat exchanger with the processor. Everything else you can build around this. Disspiating some 100 watt of power doesn't seem too hard
>That said. Laptops are plenty powerful already. I have a 4-5 year old laptop, with an 8 core processor. I can run pretty heavy code on it.
>>4548
>For my personal use, a weak machine is more than enough.
Laptops these days are more powerful than anyone needs
>For everything computationally intensive, makes way more sense to use a desktop. Idk why desktops lost favour with people.
normalfags think what the tv tells them to. And no one ever advertises a cool guy doing cool things on a desktop, it's always a laptop or a tablet or some other weird gadget so normalfags associate laptops with being cool and hip and desktops not so. There. Apart from this a laptop is more personal and portable and unlike 8chan rahiwasis like ourselves people often step out and work elsewhere, so I suppose that is something.
also it's obvious that laptops are harder to repair and easier to convince people theirs is outdated.
if I have enough space and don't need the portability I will use a desktop
>>4552
>the reason why M series apple chips are so good is because the RAM is on-die instead of being separate and they put in a fuckton of accelerators
How does RAM being on die make a computer faster? (I am asking because I don't know kek)
> if you bring in RISC CPUs, not ARM only, then PPC macs, SPARC workstations and all sorts of high-perf workstation machines have existed for as long as x86 has
if you have more resources about this link them this seems interesting