Why is Apple's M1 so fast?

pagan magick

>>1
1.Larger L1 cache
2.Less cruft in Armv8 design vs x86
3.More power efficiency due 4+4 bigLittle design
4.5nm chip vs 14nm intel/7nm amd.

S

>>1
It's not.

>>4
It is. Verify your benchmarks.

Dependent types would have prevented this.

Bloat hasn't caught up yet

Wish I could've seen the look on Cudder's face when he learned that CISC is dead and x86 is destined for the trashcan.

>>8
Relying on architecture specific ASM is an anti-pattern that
killed tons of software projects that though their special, specific machine architecture will stand the test of time:
The actual lifespan of most architecture is limited by design, since
adding shit on top of old architecture(x86) eventually hits limits and performance gaps emerge which are later exploited by new cruft-free designs.

>>8,9
lol no

This is what happens when a single company controls everything from software to hardware.

Even if it's faster, Apple's proprietary locked-down anti-user shit is not worth having anyway.

Do any of you have definitive proof that Mossad is not responsible?!

My burning Yule log makes it go slower, wtf?

>>10

You're missing the point. Forget about Apple; what's to prevent other manufacturers from doing the same? Build lighting fast ARM hardware, save shitloads of power and leave x86 compatibility for virtual machines to deal with. RIP Intel

>>10
That's called "system engineering" and the premium price is what you pay for that level of hard&software integration. Sun and SGI were far better at it, but Apple is about all we have now.

The only relevant thing about M1 is less power consumption. CPUs already are insanely fast but software is slow anyway.

>>13
x86 emulating ARM adds an order of magnitude slowdown, similar for ARM emulating x86. It might be possible to make something that is natively as fast by reverse engineering x86 microcode and having it implement both x86 and ARM. This would be quite a feat especially given the complexity of both architectures when you consider all the multimedia, systems, virtualization and other extensions.

RIP intel

https://www.tomsguide.com/amp/news/microsoft-reportedly-creating-own-arm-chips-to-fight-macbook-m1

Microsoft's reason to develop non-Intel chips is to exclude Free operating systems from booting on bare hardware. They're fine with GNU/Linux as a VM client as long as their partners' ads still run in the bottom right corner.

>>18
search "microsoft pluton"

>>19
Allow me to revise my previous comment: ``Microsoft's reason to develop Trusted Platform shite is, and has always been, to exclude Free operating systems from booting on bare hardware.''

With Secure Boot it was about supposedly invisible bootloader-staged hypervisor malware, which doesn't exist in the wild but the signed-bootloader shenanigans Microsoft got up to were certainly enough to prevent users from running kernels they compile themselves. Secure Boot is mandatory on Microsoft-approved ARM hardware, and with ARM there's no Linux Foundation demanding that the bootloader signature verifier recognize root keys installed by the hardware owner.