Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: Why hasn't x86 caught up with Apple M series?

news.ycombinator.com

Post reply on HN
Print view View original

561–570 of 640 posts

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#561

Post by ben-schaaf »

alt227 wrote: post #491
alt227 wrote: post #364

Earlier quoted context omitted.

> A huge reason for the low power usage is the iPhone. No, the main reason for better battery life is the RISC architecture. PC on ARM architecture has the same gains.

Any downvoters care to actually leave me a reply telling me why? Im not wrong!

You are wrong. The Snapdragon X Elite is actually a great example, unlike M1 it's performance isn't particularly great and it eats 50W under load. That makes its CPU cores a fair bit less efficient that AMDs even on the same production node. If Apple Silicon didn't exist then you might instead argue that x86-64 is more efficient than ARM.

If all that's true then why does Snapdragon have better battery life? As I said in my comment the great battery life comes from when the CPU isn't being used. It's everything else around it. That's where AMD is still significantly behind.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#562

Post by tacticalturtle »

alt227 wrote: post #491
alt227 wrote: post #364

Earlier quoted context omitted.

> A huge reason for the low power usage is the iPhone. No, the main reason for better battery life is the RISC architecture. PC on ARM architecture has the same gains.

Any downvoters care to actually leave me a reply telling me why? Im not wrong!

You might find these posts informative:

https://chipsandcheese.com/p/arm-or-x86-isa-doesnt-matter

https://chipsandcheese.com/p/why-x86-doesnt-need-to-die

All instructions across x86 and Arm are being decoded to micro-operations, which are implementation specific. You could have an implementation which prioritizes performance, or an implementation that prioritizes power consumption, regardless of the ISA.

Decoding instructions, particularly on a modern die, doesn’t consume a significant amount of area or power, even for complicated variable length instructions.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#563

Post by ben-schaaf »

RajT88 wrote: post #366
ben-schaaf wrote: post #101

Battery efficiency comes from a million little optimizations in the technology stack, most of which comes down to using the CPU as little as possible. As such the instruction set architecture and process node aren't usually that important when it comes to your battery life. If you fully load the CPU and calculate how much energy a AI340 needs to perform a fixed workload and compare that to a M1 you'll probably find s…

> All that to say: M1 is pretty fast, but the reason the battery life is better has to do with everything other than the CPU cores. That's what AMD and Intel are missing. Apple is vertically integrated and can optimize at the OS and for many applications they ship with the device. Compare that to how many cooks are in the kitchen in Wintel land. Perfect example is trying to get to the bottom of why your windows lapto…

This is easy to disprove. The Snapdragon X Elite has significantly better battery life than what AMD or Intel offer, and yet it's got the same number of cooks in the kitchen.

> Perfect example is trying to get to the bottom of why your windows laptop won't go to sleep and cooks itself in your backpack

Same thing happens in Apple land: https://news.ycombinator.com/item?id=44745897. My Framework 16 hasn't had this issue, although the battery does deplete slowly due to shitty modern standby.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#564

Post by rerdavies »

fluoridation wrote: post #465
hajile wrote: post #451

Earlier quoted context omitted.

Intel has proposed APX to address this. It does away with some of the 32-bit garbage that complicates design for no good payoff. Most importantly, it increases from 16 to 32 registers and allows 3-register instructions (almost all x86 instructions are 1-register or 2-register instructions). This would strip out tons of MOV instructions which was proven with AMD64 to have a decent impact on performance. Reduced I-Cach…

>This would strip out tons of MOV instructions which was proven with AMD64 to have a decent impact on performance. Sure, but the topic is optimizing power efficiency by removing support for an instruction set. That aside, if an instruction isn't very performant, it isn't much of an issue per se. It just means it won't get used much and so chip design resources will be suboptimally allocated. That's a problem for Inte…

ARM has THREE instruction sets. (four?) aarch32, aarch64, and various incarnations of Thumb. (A PI 5 supports all three).

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#565

Post by ben-schaaf »

pzo wrote: post #142
ben-schaaf wrote: post #101

Battery efficiency comes from a million little optimizations in the technology stack, most of which comes down to using the CPU as little as possible. As such the instruction set architecture and process node aren't usually that important when it comes to your battery life. If you fully load the CPU and calculate how much energy a AI340 needs to perform a fixed workload and compare that to a M1 you'll probably find s…

> most of which comes down to using the CPU as little as possible. it least on mobile platform apple advocate the other way with race to sleep - do calculation as fast as you can with powerful cores so that whole chip can go back to sleep earlier and more often take naps.

Race to sleep is all about using the CPU as little as possible. Given that the modern AMD chips are faster than Apple M1 this clearly does not account for the disparity in battery life.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#566

Post by ben-schaaf »

mrtksn wrote: post #244
ben-schaaf wrote: post #101

Battery efficiency comes from a million little optimizations in the technology stack, most of which comes down to using the CPU as little as possible. As such the instruction set architecture and process node aren't usually that important when it comes to your battery life. If you fully load the CPU and calculate how much energy a AI340 needs to perform a fixed workload and compare that to a M1 you'll probably find s…

Which also should mean that using that M1 machine with Linux will have Intel/AMD like experience, not the M1 with macOS experience.

Yes and no. The optimizations made for battery life are a combination of software and hardware. You'll get bad battery life on an M1 with Linux when watching youtube without hardware acceleration, but if you're just idling (and if Linux idles properly) then it should be similar to macOS.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#567

Post by rerdavies »

BirAdam wrote: post #250

First, Apple did an excellent job optimizing their software stack for their hardware. This is something that few companies have the ability to do as they target a wide array of hardware. This is even more impressive given the scale of Apple's hardware. The same kernel runs on a Watch and a Mac Studio. Second, the x86 platform has a lot of legacy, and each operation on x86 is translated from an x86 instruction into RI…

The "RISC" thing is an experiment that failed in the 90s. There's nothing particularly RISCy about the ARM instruction set. It is a pretty darned complicated instruction set.

ARM processors ALSO decode instructions to micro-ops. And Apple chips do too. Pretty much a draw. The first stage in the execution pipelines of all modern processors is a a decode stage.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#568

Post by rerdavies »

mxey wrote: post #421
alt227 wrote: post #358

Earlier quoted context omitted.

Apple purposely make it so after 3 new versions of the OS you cannot upgrade the OS on the hardware any further. This in turn means you cannot install new software as the applications themselves require the newer versions of the OS. It has been this way on apple hardware for decades, and has laid the foundation of not ever needing to provide backwards compatibility for more than a few years as well as forcing new har…

> Apple purposely make it so after 3 new versions of the OS you cannot upgrade the OS on the hardware any further. A new major version of macOS comes out every year. The oldest Mac still supported by the upcoming macOS 26 is from 2019.

Depends on the OS version. I had a Mac Mini blown away after three years.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#569

Post by rerdavies »

olejorgenb wrote: post #352
stego-tech wrote: post #319

There’s a number of reasons, all of which in concert create the appearance of a performance gap between the two: * Apple has had decades optimizing its software and hardware stacks to the demands of its majority users, whereas Intel and AMD have to optimize for a much broader scope of use cases. * Apple was willing to throw out legacy support on a regular basis. Intel and AMD, by comparison, are still expected to run…

> Apple has had decades optimizing its software and hardware stacks to the demands of its majority users, whereas Intel and AMD have to optimize for a much broader scope of use cases. But as you mention - they've at multiple times changed the underlying architecture, which surely would render å large part of prior optimizations obsolete? > Software in x86 world is not optimized, broadly, because it doesn’t have to be…

> Do ARM software need optimization more than x86?

If Windows-World software developers could one day announce that they will only support Intel Gen 14 or later (and not bother with AMD at all), and only support the latest and greatest NVidia GPUs (and only GPUs that cost $900 or more), I'm pretty sure they would optimize their code differently, and would sometimes get dramatic performance improvements.

It's not so much that ARM needs optimizations more, but that x86 software can't practically be broadly optimized.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#570

Post by rerdavies »

jayd16 wrote: post #360
stego-tech wrote: post #319

There’s a number of reasons, all of which in concert create the appearance of a performance gap between the two: * Apple has had decades optimizing its software and hardware stacks to the demands of its majority users, whereas Intel and AMD have to optimize for a much broader scope of use cases. * Apple was willing to throw out legacy support on a regular basis. Intel and AMD, by comparison, are still expected to run…

I generally agree but what's Qualcomm's excuse?

Samsung seems to have some ARM processors that compete favorably with M-class processors.
Post reply on HN