Apple M2 vs. AMD Rembrandt vs. Intel Alder Lake Linux Benchmarks

[tunnelblick]

Count me in on the side of people that say the fact that the MBA is fan-less is enough to show how much better Apple silicon actually is. There is just no comparison on the x86-side of things.
For a benchmark it seems this is sort of click-baity as I also think it would have made more sense to run the tests on an MBP but it's Michael's playground here so it is what it is.
I think for the current state of Asahi this is still impressive overall while probably not convincing many people to buy an Apple laptop just to use Linux on it but that's fine, not everybody is Linus Torvalds, Sven Axboe or Hector Martin.
I think it is fair to assume, given the engagement this article has, Michael will run more benchmarks on Apple silicon in the future.

[Dukenukemx]

Count me in on the side of people that say the fact that the MBA is fan-less is enough to show how much better Apple silicon actually is. There is just no comparison on the x86-side of things.
For a benchmark it seems this is sort of click-baity as I also think it would have made more sense to run the tests on an MBP but it's Michael's playground here so it is what it is.

I like how people praise Apple for not using a fan then complain when the benchmarks were done on a M2 without a fan. You can't praise Apple for not needing a fan but then point out how you needed to test it with a fan.

I think for the current state of Asahi this is still impressive overall while probably not convincing many people to buy an Apple laptop just to use Linux on it but that's fine, not everybody is Linus Torvalds, Sven Axboe or Hector Martin.

The Apple corporate trolls are fun. Yes go buy an Apple where you can't replace the SSD and randomly for no reason it sends 13V to your NAND and kills the SSD. Because at some point Linus Torvalds has used a Mac. Hector Martin is porting Linux to Mac because Mac OSX is so good that he's replacing it with Linux. Maybe in 4 years you too can run glxgears and be amazed at Apple's genius. Or buy a real PC and actually use it now with the plethora of Linux distros to choose from.

[tunnelblick]

I like how people praise Apple for not using a fan then complain when the benchmarks were done on a M2 without a fan. You can't praise Apple for not needing a fan but then point out how you needed to test it with a fan.

I didn't say that, though. But what to expect when someone starts a comment with "I like how", tells you about their style of arguing.
But perhaps you can point us to fan-less laptops from x86 that are in the same ballpark of performance.

The Apple corporate trolls are fun. Yes go buy an Apple where you can't replace the SSD and randomly for no reason it sends 13V to your NAND and kills the SSD. Because at some point Linus Torvalds has used a Mac. Hector Martin is porting Linux to Mac because Mac OSX is so good that he's replacing it with Linux. Maybe in 4 years you too can run glxgears and be amazed at Apple's genius. Or buy a real PC and actually use it now with the plethora of Linux distros to choose from.

Are you accusing me of being an Apple corporate troll? What is the point of that even?
I guess it is time to put you on ignore.

[Raka555]

I can say this:


c) M1/2 has an considerable advantage in low power single threaded workloads - I can give you that LOL

The single most important metric on a computer ...

[Raka555]

but do you know what you could get outside of Apple for that much? A whole lot more laptop.

The audience that buy MacBook Air will not be interested in a steaming hot laptop that sounds like it is ready for take-off and that would probably not last an hour on battery power.

[Raka555]

Apple is at fault for not putting a fan in their products. They did this with Intel Macbooks as well. Well they did but not on the heatsink.

At fault for producing a new class of laptop that nobody else is capable of ?

[Anux]

I'm pretty sure that if the benchmarks were run with the performance governor, than the efficiency gains would naturally be even greater.

For sure, the important thing here was the power consumption we can estimate for Ryzen 6850 wich was 18W at that setting if I remeber correctly. Thats roughly the same as the M2 while beeing faster than a M2 (I think the whole system is still more powerhungry than an M2 which can do 20 - 25 W from the wall).

[Paradigm Shifter]

Michael; the first page loads, but attempting to load any other page in this article the site just part loads (but loads the ads successfully) and then hangs. Tried Brave, Firefox and Chrome, with my normal profiles and with fresh ones (so no addons at all). The following screenshot is after I've closed a fullscreen ad and waited a few minutes.

broken.png

All other articles on the site load OK.

Also, when I try to "@" Michael, it "@"s user 26131 rather than user 1...?

[MadCatX]

I like how people praise Apple for not using a fan then complain when the benchmarks were done on a M2 without a fan. You can't praise Apple for not needing a fan but then point out how you needed to test it with a fan.

That's not the argument people are making. People point out that a fanless laptop with much lower power budget has a very competitive performance against machines that require active cooling to provide the same performance. You may dislike Apple however you want but the M1 is a commendable leap forward in terms of power efficiency. Unless there is a matching alternative available by that time, I'll gladly get a M-based laptop to replace my ThinkPad once Asahi Linux becomes usable for day to day workloads...

[drakonas777]

The single most important metric on a computer ...

Not really. These days even simple loads like "office work" (web, office suite, business apps etc.) can easily utilize at least few cores. Of course, having good low power efficiency allows to save battery during idle states, but this is not only because of ARM, but also because of a good software optimization regarding PM. In fact, I'd argue that the latter is far more important. We can do a simple test: lets compare side by side M1/2 running vanilla MacOS and Asahi linux regarding battery - I'll see how the second one will compare, since Linux has a superb PM, especially regarding platform/bus/peripherals, drivers overall LOL

The mere fact that hammering M1/2 with intensive multithreaded load significantly increases power usage (to the x86 U series levels basically) means that majority of that comes from data movement across RAM, cores, caches, fabric etc. Its not the actual core computation done inside ARM core who consumes that. Obviuosly M1/2 overall is much more efficient than current x86 SoCs, because there is a series of factors contributing to that: N5, SoC design, accelerators, software, ARM decoding. Of course all ARM fanboys out there think it's because of ARM only, cause ARM is magical LOL

To be honest, efficiency is the last reason why Apple have chosen ARM. Priorities were like:
a) Everything else in Apple ecosystem already was ARM based, so it was a natural and organic decision to unify architecture across the board so optimize the effort supporting two entirely different ISAs; That's an umbrella under which are like 80% of the reasons they went ARM;
b) Practically there were no worthy alternatives to ARM for them. x86 has licensing problem, RISC-V is still in it's infancy, POWER - perhaps, but ecosystem is small, as the market trends, MIPS is dead.
c) Somewhat better ARM efficiency and scaling. I'd say that's about it.

In theory, they could absolutely do a modern x86/64 uArch (which would not be optimized to a minimal transistor count/die space and high frequency as fuck like Intel and AMD do), cut down legacy shit (because they control a platform), and have almost the same level of core efficiency as they have with ARM. They absolutely could do it, it's engineering problem, not the "magic metaphysics of ARM being multitudes better than x86" LOL