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

Apple Silicon M1 Disruption

mondaynote.com

Post reply on HN
Print view View original

131–140 of 201 posts

Re: Apple Silicon M1 Disruption

#131

Post by alwillis »

jeffbee wrote: post #30
naderkhalil wrote: post #11

> Will Microsoft now offer an official Windows 10 on ARM that can run on M1 Macs?...This could create trouble in the x86 PC world. Or will Microsoft, less committed to Windows than it is to its applications running everywhere, be content to see Office and more run (and generate revenue) on M1 Macs…and how will HP, Asus, Dell and others react? This is just the beginning of a new competitive layout. Could this be good…

I think it will benefit Linux but not that way. It will force the x86 vendors to adopt the microarchitectural aspects of the M1 that make it fast in practice, instead of spending so much area/power budget chasing SPEC benchmarks. We will get better chips through competition.

It will force the x86 vendors to adopt the microarchitectural aspects of the M1 that make it fast in practice, instead of spending so much area/power budget chasing SPEC benchmarks. We will get better chips through competition.

There's only so much you can do with a 40-year old instruction set. There's certain things neither ARM or Intel can do compared to the M1.

Example: the max number of instruction decoders Intel and AMD can use is 4 and that brings tons of complexity due to x86 instructions ranging in size from 1-15 bytes. There's no way for the processor to know where the instruction boundaries lie except through trial and error essentially.

All ARM instructions are one size, so decoding them and having many instructions in flight at any given time is much easier. The M1 has 8 instruction decoders.

Short answer: the M1 can process twice as many instructions per clock cycle than Intel/AMD processors can.

The M1 and the Zen3 run neck and neck, with the Zen3 being slightly faster in some benchmarks.

But the Zen3 runs at 5 GHz and the M1 runs at 3.2 GHz, meaning if the M1 ran at 5 GHz, it would blow the Zen3 away. And of course, on a performance per watt basis, the M1 is the obvious winner.

This article gets into all the reasons why Intel and AMD are between a rock and a hard place when it comes to competing with the M1:

"Why Is Apple’s M1 Chip So Fast?"—https://debugger.medium.com/why-is-apples-m1-chip-so-fast-32...

Re: Apple Silicon M1 Disruption

#132

Post by alwillis »

mamp wrote: post #21
tyingq wrote: post #10

I do give Apple credit for a terrific design. It would be interesting to know how much of the performance is attributable to the smaller process size, on-package memory, etc. Things others will be able to replicate. It's not clear to me how much of it is true competitive advantage.

This article gives a pretty good overview of the components that make the M1 different (nb. I'm not a hardware person): https://debugger.medium.com/why-is-apples-m1-chip-so-fast-32...

This is a great article.

How about lets start with the M1 being able to process double the number of instructions per clock cycle…

Every ARM instruction is the same size; Intel instructions range in size from 1 to 15 bytes, making decoding them a pain in the ass.

Re: Apple Silicon M1 Disruption

#133

Post by stephencanon »

jeffbee wrote: post #65
kergonath wrote: post #44

Earlier quoted context omitted.

I don’t think these goals are exclusive. I am not sure about the M1, but the A13 and A14 are doing quite well in SPEC benchmarks. If they have compromised these scores, it’s not obvious to me.

I don't think they've compromised on SPEC performance at Apple, but I think they've chased it too hard at Intel. I'm sure I'd rather have more iTLB entries and more L1i cache than that third 512x512 FMA unit, if that's a trade that can be made.

Second, not third, and chasing benchmark scores (instead of real workloads) would result in exactly the trade off you’re asking for; most pieces of SPEC will barely benefit from the second AVX512 FMA at all.

Re: Apple Silicon M1 Disruption

#134

Post by icedchai »

grecy wrote: post #64
untog wrote: post #46

Earlier quoted context omitted.

> Appreciating the iPhone is largely done in hindsight. I’m sorry but that’s definitely not true. Yes, there were a vocal minority of folks who said the iPhone was nothing special but the mass market response was huge and it was heralded as a huge deal by the vast majority. And to go back to my original point: there’s a difference between “this new invention allows you to do new things, but those things aren’t import…

Did you forget when the CEO of Microsoft laughed it off and said it was crap? [1] [1] https://www.youtube.com/watch?v=eywi0h_Y5_U

To be fair, if you remember 2007, the first generation was lacking in many areas. You couldn't send MMS messages. Receiving MMS was awkward and was done through an AT&T web site. You couldn't install apps. You couldn't take videos, only photos. It was slow, etc.

My first iPhone was a 3G, but felt the iPhone really started taking off with the iPhone 4. It felt like a huge leap.

Re: Apple Silicon M1 Disruption

#135

Post by faitswulff »

gameswithgo wrote: post #106
faitswulff wrote: post #42

Earlier quoted context omitted.

I think the reason it keeps coming up is that it is the most prominent bad review, but due to Moorhead’s allegiances, poor calls in the past, and eclectic selection of software, his negative review comes across as being in bad faith.

or maybe it just really is in bad faith?

Almost certainly.

Re: Apple Silicon M1 Disruption

#137

Post by doggydogs94 »

I am looking at two compatibility points regarding M1-macOS. First, will VMWare (and Parallels and virtualBox) run x86 Windows VMs on M1 and second, will Linux run natively on M1 (even though I don’t do much Linux). Until these two things are taken care of, M1-MacOS will just be (an important) niche system.

Re: Apple Silicon M1 Disruption

#138

Post by QuarterReptile »

reader_mode wrote: post #87
jaimex2 wrote: post #59

Its only disruption if it disrupts. No one cares other than existing Apple users so I guess the only disruption will be the year of really fast and efficient teething issues they will be dealing with. We have one at work, I'd highly recommend waiting another year before getting one if you want to be productive.

If this was just a new CPU option with a performance boost over Intel (and significantly better thermals) but a compatibility mess - I would agree. But MBP is 500$ price drop from the Intel variant. At these price points these laptops are a far better value proposition than the competition - Intel thermal performance and power consumption is just terrible, AMD is better but not present in anything premium - in terms…

AMD is in Surface laptop 3, and in Thinkpad T14 (supposedly, though the ship times are reputedly several months.) The T14 is premium in a certain class (high value placed on swapping parts.) I suspect there are a few more.

Re: Apple Silicon M1 Disruption

#139

Post by AtlasBarfed »

A credible ARM response will probably come in about 1-2 years from Intel/AMD/Qualcomm/etc.

This is the second major opportunity for Linux to make an actual push onto the desktop.

Windows 8/10's abhorrent design was left at the roadside, but arguably there was no motivation from big backers to get Linux up to par on desktop.

This time may be different.

Why would AMD/Intel/Qualcomm wait/hope/gamble that Microsoft executes the transition? Especially given the major shortcomings in UI? The architecture transition will destroy Microsoft's compatibility advantage.

Why not hedge with a really well supported Linux Desktop alternative they personally/collectively fund and push... personally I think that's corporate malfeasance given the scale of revenue to not pursue. Or a BeOS.

The architecture switch will probably limit the hardware that needs support. A linux-based or at least OSS_based x86 emulation somewhat on par with Rosetta will be a critical component as well.

Re: Apple Silicon M1 Disruption

#140

Post by jeffbee »

alwillis wrote: post #131
jeffbee wrote: post #30

Earlier quoted context omitted.

I think it will benefit Linux but not that way. It will force the x86 vendors to adopt the microarchitectural aspects of the M1 that make it fast in practice, instead of spending so much area/power budget chasing SPEC benchmarks. We will get better chips through competition.

It will force the x86 vendors to adopt the microarchitectural aspects of the M1 that make it fast in practice, instead of spending so much area/power budget chasing SPEC benchmarks. We will get better chips through competition. There's only so much you can do with a 40-year old instruction set. There's certain things neither ARM or Intel can do compared to the M1. Example: the max number of instruction decoders Intel…

It sounds like you are just regurgitating some blog post you found. I don't have any workloads that are decode starved on x86 so it wouldn't make any difference if Intel had 50 decoders. I do have tons of workloads that spend a lot of time on TLB misses. I have several that are cache fill limited.
Post reply on HN