Hacker News Reader

Comment by Rebelgecko | Hacker News Reader

It's bizarre to me they're still bragging about performance relative to Intel macs. I guess there's been diminishing returns after the ARM transition, or else they'd be comparing to M1?

View on HN

For an M1 user, there's not a lot of incentive to upgrade because they already got the huge boost in terms of battery life and performance. If you love the styling, or you need a larger size, you might upgrade. But lots of people are still running Intel MBAs or MBPs and so they're pitching to that audience.

They could probably also cite stats of how it compares to the M1, but that would lead to potential confusion as journalists and bloggers mix up which stat is which. Marketing people want one message, said clearly and repeated often.

I suspect the majority of mac owners still have intel based macs so they want to highlight the performance improvement to those users - since they may be most likely to upgrade?

It's correct that there are Intel chips that last as long as ARM macs.

However, AFAIK there are none that compare to M1/M2 in terms of efficiency and relative performance.

I might be wrong - I'm comparing an old budget ThinkPad (L580) to 2021 and 2023 MacBook Pros here.

Noteworthy for me is that this 2019 i5 with 16GB (Lenovo L580) still trashes all AMD notebooks that I ever used in that regard. Xubuntu lasted me a work day easily on that machine. But performance was slow of course.

Even in single thread tasks in interpreted languages, the speed difference between that 2019 i5 and an M1 Pro is around factor 5-10, with the Mac inaudible compared to the i5 sounding like a small airplane (cough, legacy webpack frontend build).

Also here, my experiences are probably distorted by price range differences.

Then there's all the stuff that really matters: virtually unlimited multitasking, reliability (wireless, graphics especially), trackpad, speakers, WebRTC, etc etc

And with a current MBP, you really need to do very expensive tasks (ML, huge builds etc) to notice any performance or battery life degradation.

I had a 13" MBP with the last Intel chips and no whatever bar. I got at most 4 hours of battery life during the work day.

I now have a 2022 M2 13" and I can go 10+ hours w/o a charge doing normal work shit which generally includes 4-6 hours of video chat and 2-4 hours of development.

For me, the upgrade was well worth it. I now treat my computer like I do my phone; I charge it once a day overnight.

10h would be nice. Depends on workload significantly I bet; any x86 emulation likely tanks it. 6h of video processing is seriously impressive though.

On Apple's own site they claim "up to" 40% faster than M1 Air. Maybe that's true? The problem with comparing it to contemporary Intel models is neither their performance nor efficiency looks terribly amazing. They are marketing to existing macOS users and it is easy to demonstrate that M2 stomps all over a 10th-generation Core MBA.

I think it could be a combination of: 1. Marketing speak because it sounds better to compare it to the much slower Intel Macs than to the fast M1. Plus the gains between and M1 and M2 are marginal compared to Apple Silicon vs. Intel. 2. They are marketing more to Intel Mac owners/potential PC switchers rather than M1 owners.

The chip is the same as the last year’s, only the screen size is larger.

M2 vs M1 is a minor difference (relatively). It feels like M3 was / is delayed so there's nothing better to brag about.

Note: maybe not M3 the name, but at least the 3nm transition.

It's useful for folks currently running Intel Macs (or considering a used one). For other folks, not as useful since the last Intel MacBook only ran 10th gen. As an example, the top end 13 inch 2020 MacBook Pro used an i7-1068NG7. A current Intel i7-1360p will get about 2.4x the multi-core performance of that old 10th gen chip, even outperforming an M2 in most multi-core workloads.

Is it a marketing trick to make people think they're that much faster than current Intel computers?

Depends heavily on the workload. A lot of tasks can use the newer accelerators in the M1/M2 over what was in 10th gen Intel. The higher end Intel models also had the issue of running hot and stuttering in practice which really sucked, as underclock/undervolt would have made the experience much better even at the cost of 5% performance on all-out workloads.

That doesn't get into the battery life. My M1 air has mostly been used on road trips to check email, or do light reading. No video or work. I can get pretty much a full 4-6 day trip on a full charge just doing email/reading in the evening a couple hours a night. This has been pretty fantastic imo. Never had any other device that could achieve that before.

I've used M1 Max and M2 Pro for work, and both are fantastic... x86_64 docker used to be a pretty big issue for some images, but it's been great outside of that. The drive I/O is outstanding, and can build large web/node apps faster than my 5950X desktop with a 980 pro drive.

Biggest down sides come down to the money sink if you want/need more storage or memory in the box. It's WAY overpriced for what it is, and you might be better served with a DIY desktop or framework laptop. There are also some workloads that are a true miss. Not to mention gaming as an entire use case that is pretty poor. Not to mention the UX is rather dated at this point, and I really wish some of the hotkeys were more aligned with Windows and Linux. Muscle memory is a pain when you jump back and forth.

I think the major advantage is performance per watt and low heat, rather than raw performance?

Airs don't even need a fan.

Weird way to say it. Airs are underclocked to avoid reaching fan temps.

Maybe they are making that comparison for the people they want to upgrade to the new macs and not for people that just got their M1 MacBooks.. making m1 to m2 comparison for intel mac users would not make much sense in that case.

There is still a lot of people to move over to the ARM side. Most people haven't made the move yet. I'm just about to myself. It makes more sense for them to focus on migrating people over at this stage.

I think it's also a clever way to compare performance to other, non-Apple Intel machines--without mentioning them.

Its good. It just means they are not looking for M1 users to upgrade.

Reply on news.ycombinator.com