Will X86 ever die? I know backwards compatibility etc. etc. but

"no"

ARM shit is only on crappy phones, and you can't do any productive work on them without wanting to kill urself so i don't see this happening any time soon.

>>56349418
>ARM is a measurably better architecture
It is in theory but intel is lightyears ahead of ARM in terms of performance per watt and IPC.

Anyway x86 may never die because not only does it support old stuff but it's the best cpu architecture for enterprise and supercomputers.

>>56349435
> x86 is the best cpu architecture for supercomputers.

k.
Not like power8 exists or anything.

Also, arm uses a limited instruction set. While x86 uses a full featured set. Both have their strengths and weaknesses...

>>56349418
Also, depends. higher level languages will be fine, but lower levels will not. For example, with assembly, you would have to re-write programs to deal with issues like less registers. You also might have to substitute specific instructions with groups of instructions going from one architecture to the other

>It wouldn't be too hard to recompile x86 programs into ARM anyways right?

Anon, a reasonable number of programs still in use are abandoned. They will never be recompiled.

>>56349543
>Not like power8 exists or anything.
It's irrelevant, just a meme architecture IBM designs with the left over money they have.

>Also, arm uses a limited instruction set. While x86 uses a full featured set. Both have their strengths and weaknesses...
If you're talking about the RISC vs CISC thing then it means jack shit now. Modern X86 processors incorporate a RISC core.

Anyway let me know when ARM can reach i7 multi-core performance with just 45W max TDP and only 8 cores.

http://www.cpubenchmark.net/compare.php?cmp%5B%5D=2718&cmp%5B%5D=2565

x86 is literally the foundation of computing. How can we even have other architectures if we aren't using x86 as a base to build them on? x86 will never go out of style. It's a fixture and nothing can displace it. It will always be the king of its niche (computers).

>>56349418
>X86 ever die

Why would it? Nobody is just going to reinvent something like that when it's not broken and there's billions of man hours of tooling already done for it.

>>56349763

Because it's already "terrible" in modern applications. It has two things going for it number one, it works, and number two it runs everything reasonably fast.

We likely won't see anything to replace it for a very long time. Not until we are deep in to molecular sized CPU's.

>>56349763
>>56349637
Its pretty weird when you think that x86 is only the standard for computers because IBM chose to put an intel 8088 in the IBM PC

>>56349835
>Because it's already "terrible" in modern applications.
Can you elaborate because >>56349435 is right.

>>56349835
>Because it's already "terrible" in modern applications

Literally the fuck are you talking about?

>>56349844
Not really. There doesn't need to be more than one standard since any relatively advanced architecture is more or less interchangeable, and there's a lot of benefit in picking one so you don't have a lot of redundant work being done.

>>56349418
Everyone and their mother has tried to replace x86 for over 20 years. Wanna guess how successful they have been?

>>56349418
Well, uh, ARM puts a lot more strain on the memory and cache infrastructure due to the smaller instruction set. To be competitive with x86 it'd need even more cache than this bloated mess does.

>>56349418
Yes, when everything moves to the cloud over the next 10 years or so. With Moore's law being dead and backwards compatibility no longer being as much of an issue it should be much easier for a new architecture to step in to increase performance.