M1 Macs: Truth and Truthiness • Hacker News Board

Re: M1 Macs: Truth and Truthiness

#721

Post by alwillis » Fri, Dec 04, 2020, 4:45 AM UTC

ianhowson wrote: post #540

alwillis wrote: post #501

Earlier quoted context omitted.

>M1 is using standard LPDDRx. >It's not "very high performance". AnandTech disagrees with you: https://www.anandtech.com/show/16252/mac-mini-apple-m1-teste... Besides the additional cores on the part of the CPUs and GPU, one main performance factor of the M1 that differs from the A14 is the fact that’s it’s running on a 128-bit memory bus rather than the mobile 64-bit bus. Across 8x 16-bit memory channels and at LPDD…

Anandtech is comparing M1 vs A14. It's high performance for a cellphone part. Dual channel DDR3L or DDR4L also has a 128 bit bus. 4200MHz DDR4 is clocked on the high side for most laptops, sure, but it's hardly unusual. Run the numbers and you get the exact same throughput figure as for M1, which isn't surprising, because we're just taking width * rate = throughput. So I'll repeat my assertion, downvotes be damned: t…

Anandtech is comparing M1 vs A14. It's high performance for a cellphone part.

That’s where they started, but their conclusion was beyond that.

Did you miss the part where they said the fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before?

This isn’t only about A14 vs M1.

It’s not that LPDDR4X-4266-class memory is special; it’s been around for a while. What is special is that the RAM is part of SoC and due to the unified memory model, the CPU, GPU, Neural Engine and the other units have very fast access to the memory.

This is common for tablets and smartphones; it’s not common for general purpose laptops and desktops. And while Intel and AMD have added more functionality to their processors, they don’t have everything that’s part of the M1 system on a chip:

* Image Processing Unit (ISP) * Digital Signal Processor (DSP) * 16 core Neural Processing Unit (NPU) * Video encoder/decoder * Secure Enclave

There’s no other desktop such as the M1 Mac mini that combines all of these features with this level of performance at the price point of $699.

That is special.

Re: M1 Macs: Truth and Truthiness

#722

Post by ogre_codes » Fri, Dec 04, 2020, 4:46 AM UTC

ntsplnkv2 wrote: post #690

ogre_codes wrote: post #411

Earlier quoted context omitted.

> Last year it was just that Apple had some easily replicated It feels like if you listen to the peanut gallery, Apple CPU advantage is always easy to replicate. They said the same thing when Apple bumped the A series to 64 bits and at pretty much every step between. If Apple's formula was easy to copy or as obvious as everyone seems to claim, then we wouldn't be in the position where Apple has the best mobile CPU, t…

Apple always gets people riled up on here, it's one of the few areas on HN that discussions just get rather poor quickly. > If Apple's formula was easy to copy or as obvious as everyone seems to claim, then we wouldn't be in the position where Apple has the best mobile CPU, the best wearable CPU, the best tablet CPU, and the best ultrabook CPU. Exactly. It isn't easy to do. Apple is in a unique position to push their…

> And while that sacrifices openness, most people buying laptops just don't care.

This is it in a nutshell. I used Linux as my web development platform because I could immerse myself in Unix on both the server and client (Something you still can't do well in Windows FWIW). When OS X moved to Intel, suddenly I could live and breath something very close to my server side Linux setup on my desktop and have a fuss-free OS. Open is nice, but performance & just being able to get my job done effectively is much more important.

Re: M1 Macs: Truth and Truthiness

#723

Post by ogre_codes » Fri, Dec 04, 2020, 4:47 AM UTC

the_af wrote: post #685

ogre_codes wrote: post #317

Earlier quoted context omitted.

I suspect people interested in "superfast, supercool ARM-based Linux laptops" are looking more for running on bare metal, not in a VM.

If Linux can be made to run on M1 bare metal, I expect it will. And computing history shows us Linux can be made to run on anything ;)

I expect (very much hope) it will. I haven't contributed to the Linux on M1 fund yet, but I am seriously considering it, particularly if I get an M1 Mac. I likely wouldn't run Linux until after Apple retires MacOS for the M1, but want the option.

Re: M1 Macs: Truth and Truthiness

#724

Post by josephcsible » Fri, Dec 04, 2020, 4:49 AM UTC

ekianjo wrote: post #623

erk__ wrote: post #303

Earlier quoted context omitted.

I am pretty sure the most open at the moment are OpenPOWER and RISC-V, while there is is MIPS I am sure how open it is.

Can you buy a desktop with those right now?

POWER: Yes, https://www.raptorcs.com/content/base/products.html

RISC-V: Not really. The closest thing is probably https://www.sifive.com/boards/hifive-unmatched

Re: M1 Macs: Truth and Truthiness

#725

Post by nwienert » Fri, Dec 04, 2020, 4:50 AM UTC

sagarm wrote: post #677

nwienert wrote: post #640

Earlier quoted context omitted.

For years we've been evolving. I've owned ~6 laptops over the last 10 years, all Intel Macs. Each year you get about ~20% ~performance increase for the same "costs" (weight, battery, price). This year you get about 100% performance increase and the costs are all halved. Ok, so... we'll call it a "huge" evolution. All I know is I've been waiting since the air came out for this laptop to exist: fast enough to handle al…

AMD laptops are also crushing this generation of Intel laptops. AnandTech's comparison shows that M1 vs AMD is mostly a power/performance trade-off at the moment. Not that the M1 isn't exciting especially in the ultrabook segment.

What's the watt / battery comparison?

Re: M1 Macs: Truth and Truthiness

#726

Post by asveikau » Fri, Dec 04, 2020, 4:58 AM UTC

innagadadavida wrote: post #604

jitl wrote: post #542

Earlier quoted context omitted.

What lockdown is grating to you? I grew up on BSD multi-user systems and Apple things still look the same to me: crusty BSD with ancient coreutils in the terminal, nice rounded corners in the GUI. What is the iPad-ification? Just the code signing stuff? I guess you can't write to /System anymore these days? And there's something about app notarization, but it doesn't really affect any of my uses.

Almost all GNU user land tools and libraries work on BSD and you can install them fairly easily with brew. Unless you are doing Linux kernel development, there is little you’ll miss out on.

I'm typing this on FreeBSD. If the most interesting part of a Mac to you is the BSD layer ... Well, I doubt that. People like Mac for the UI and Cocoa stuff, plus hardware and integration thereof.

The way you've substituted "BSD" for macOS really doesn't make sense. Yes, there is BSD code in MacOS, quite a bit of it in fact. But they're not the same thing.

Re: M1 Macs: Truth and Truthiness

#727

Post by 1123581321 » Fri, Dec 04, 2020, 4:59 AM UTC

PaulDavisThe1st wrote: post #506

__d wrote: post #489

Earlier quoted context omitted.

I'm interested to see what Apple comes out with for the higher-end MacBook Pros, iMac Pro, and Mac Pro. I expect to see the M1 used in the iMacs, although ... That could be a single higher-end chip, with possible multi-socket support for the Mac Pro? Or it could be two chips: one for the iMac and laptops, and a beefier one for the Pro desktops? Either way, they absolutely need to compete with Zen3 on highly parallel…

Laptops are almost always going to weight CPU/Watt fairly strongly. As a developer, I don't give a damn about CPU/Watt, I care only about maxCPU, for a given budget. I don't know why any company would build a laptop that focused on maxCPU. Gaming laptops are perhaps the exception, but Apple has historically not really participated in this subset. >for many devs, basic CPU core counts and high I/O speeds are the winne…

As a developer, I find long battery life, while still being able to work, to be quite nice. Power per watt is how this want is accomplished by laptop and OEM parts manufacturers.

Re: M1 Macs: Truth and Truthiness

#728

Post by sudosysgen » Fri, Dec 04, 2020, 5:00 AM UTC

threeseed wrote: post #713

sudosysgen wrote: post #706

Earlier quoted context omitted.

Similar performance and cooling? Buy a Ryzen 4800U laptop. Similar battery life? A good part of it is down to the software, sadly. For the trade-off of 8 instead of 12 hours of battery life, you'll have a repairable and open computer. Seems competitive. You also get upgradeable RAM, a GPU that is actually useful because it has drivers for a useful API, and much more I/O.

1) M1 is significantly faster than 4800U at single-core. 2) M1 is up to 20 hours of battery life. 3) Not the software since the existing Intel based Macs share the same OS.

M1 is somewhat slower in multicore, and has a fifth of the I/O performance. They are comparable.

Importantly, you're comparing the M1 in a Mac Mini, to the 4800U in a laptop. Comparing the M1 in the Air, for example, the 4800U is quite a bit faster in multicore and the single core performance advantage shrinks.

As for 2 and 3, you're comparing to AMD, not to Intel. AMD CPUs have significantly better power efficiency than Intel.

If you want to compare battery life without taking into account software, compare power consumption. The Zen chip uses slightly more power than the M1 on a laptop.

Re: M1 Macs: Truth and Truthiness

#729

Post by toast0 » Fri, Dec 04, 2020, 5:05 AM UTC

rmrfrmrf wrote: post #692

kag0 wrote: post #630

Earlier quoted context omitted.

I agree in general, but is this really true? > the legacy chip makers have a business model that doesn't work It seems like this implicitly assumes that everyone switching to Apple silicon is a possibility. But it doesn't seem likely that all the android users are going to switch to iphones or that all the windows/linux users are going to switch to macs. This is hugely embarrassing for Intel and Qualcomm in particula…

I think it has less to do with everyone switching to Apple silicon and more to do with the clarification of viable business models: you either become an IP company and outsource production (AMD) or you go completely vertical and design and produce custom-designed SoCs to be used exclusively in your consumer products (Apple). The Intel model of designing and producing all-purpose commodity CPUs seems to have hit a wal…

The (apparent) lack of demand for intel chips is because their manufacturing problems have disrupted their roadmaps. Because their 10 nm process is still not really working, and they were working on multiple releases assuming the process would be there when the design was, they're in a bad place.

From public statements, it seems like they've thought 10 nm was going to work soon for the whole time since 2018 that it's been in limited production. And they haven't really done substantial design work on the 14 nm processors (there's some releases, but it's mostly smashing more cores on there, and less architectural changes; afaik, coffe lake and comet lake don't have much in the way of ipc gains vs kaby lake). Also, they clearly need a new naming person.

It doesn't help that they killed their atom for phones program wheb they did. Right around the time Microsoft announced continuum for windows mobile 10, which would have been much cooler if it was running on an x86 chip (and if Microsoft didn't do a terrible job with Windows mobile 10 in general).

Re: M1 Macs: Truth and Truthiness

#730

Post by sudosysgen » Fri, Dec 04, 2020, 5:07 AM UTC

ENOTTY wrote: post #718

sudosysgen wrote: post #708

Earlier quoted context omitted.

> Being a process node ahead is already "much better", and IS a feat. Where are the process node ahead chips from the competitors? Not if you still don't beat them cleanly. The end result is what matters. Also, being able to pay more than other people isn't a technical feat. > On top of that, it's Apple's own SoC design, it has its own GPU design and co-processors, tons of work into making speed optimizations all aro…

> they still carry all the x86 baggage in microcode, since tons of bad decisions leading to things like Spectra. ARM doesn't > AMD is not affected by Spectre This is not true. AMD, ARM, Intel, and IBM are all affected by Spectre (Bounds Check Bypass) and that particular variant has not been fixed in hardware, and short of disabling speculative execution or radical changes to ISA or microarchitecture, is likely to nev…

AMD's new CPUs, post FX, are not affected by the majority of Spectre vulnerability. The ones that do affect them are fixed in software with essentially no performance penalty. Less than even ARM CPUs.