Most active commenters
  • gazelleeatslion(3)
  • totalZero(3)
  • manojlds(3)

←back to thread

1080 points antipaul | 62 comments | | HN request time: 0.969s | source | bottom
1. faitswulff ◴[] No.25066407[source]
This really flips the argument that Mac hardware is overpriced and underpowered on its head. Now Apple computers are a premium product from a performance perspective, as well.
replies(13): >>25066438 #>>25066460 #>>25066481 #>>25066592 #>>25066593 #>>25066611 #>>25066666 #>>25066717 #>>25066772 #>>25067074 #>>25067563 #>>25068229 #>>25071110 #
2. silly-silly ◴[] No.25066438[source]
Its not though, The 16 inch has known performance bottlenecks on both thermal and power draw.
replies(3): >>25066445 #>>25066446 #>>25066570 #
3. Diesel555 ◴[] No.25066445[source]
Source please? Honestly curious, on one now.
4. gjsman-1000 ◴[] No.25066446[source]
That's disputed. But when they get around to updating that 16-inch next year, holy smokes.
5. gazelleeatslion ◴[] No.25066460[source]
So performant premium that I’m thinking of calling Apple about my MBP preorder and switching it to the Air.

Last thing I need is a MacBook Air equivalent with an unnecessarily loud and annoying fan.

replies(3): >>25066493 #>>25066525 #>>25066526 #
6. chrysoprace ◴[] No.25066481[source]
Does it? The article is comparing Apple products, instead of comparing an Apple product to an equivalent performer from a competitor.

It's not really Apples to Apples (even if it is in name), so to speak.

replies(1): >>25066500 #
7. whalesalad ◴[] No.25066493[source]
I would. It’s the same chip in every device. Plus, having owned one of each, the chassis of the Air cannot be beat.
replies(2): >>25066505 #>>25066579 #
8. jaykru ◴[] No.25066500[source]
It's comparing Apple Silicon to Intel silicon (in an Apple product), which is apples to oranges to me.
replies(1): >>25066565 #
9. gazelleeatslion ◴[] No.25066505{3}[source]
Waiting for some more information. Fan noise on any MBP is Hell and out of control. Hoping this brings back to normal.

So happy Apple is dumping Fantel chips

10. tedk-42 ◴[] No.25066525[source]
Hang on, you're swapping a AS MBP for an AS MBA because you think the fan noise will be an issue with the MBP?

If anything, it's a bonus to have the fan so your can have prolonged boost performance while the MBA chip will throttle under continuous load.

replies(1): >>25066551 #
11. marta_morena_28 ◴[] No.25066526[source]
Yeah, they only put the fan in MBP to annoy people, of course. It has zero use. Go order the Macbook Air and have fun. Maybe you can fry your breakfast eggs on it too.
12. gazelleeatslion ◴[] No.25066551{3}[source]
Any MBP with Fantel chips (brand new, no dust) is so quick to just gun it with the fan noise - slack, chrome, electron apps, second monitor.

I think I would rather take a small performance hit and some heat than have Apple quick to pull the trigger with a fan blasting noise as I’m trying to focus (if yield comparable stats).

Will wait for more info. If this chip is really that much more efficient hopefully we are back to the good old days where MBP fan is tolerable. Otherwise I’m all in on Air

replies(1): >>25066708 #
13. arthurcolle ◴[] No.25066570[source]
Tell me more, I'm dying to go to the Apple Store and get this replaced.
replies(1): >>25066833 #
14. lostlogin ◴[] No.25066579{3}[source]
An air with 2 plugs on each side would beat it. Bonus points if MagSafe returned.
replies(1): >>25066606 #
15. ProAm ◴[] No.25066592[source]
> This really flips the argument that Mac hardware is overpriced and underpowered on its head.

This doesn't excuse years behavior not respecting their customers on (price + performance) divided by bugs.

replies(2): >>25066673 #>>25066694 #
16. michaelmrose ◴[] No.25066593[source]
Shouldn't we wait for non synthetic benchmarks to be performed by third parties running real applications?

We could even compare some cross platform apps across both OS and cpu and see how the total package performs.

replies(2): >>25067331 #>>25070415 #
17. whalesalad ◴[] No.25066606{4}[source]
I don’t miss MagSafe. A good external monitor will charge the Mac and serve as a usb/thunderbolt hub.
18. ekianjo ◴[] No.25066611[source]
So we trust Geekbench over actual user testing now?
replies(1): >>25068824 #
19. Existenceblinks ◴[] No.25066666[source]
In my perspective, the underpowered part has always been graphic card though.
20. willio58 ◴[] No.25066694[source]
Customers seem to be pretty happy with Apple in general so I can’t imagine they feel disrespected by the price and performance.
replies(1): >>25066880 #
21. kube-system ◴[] No.25066708{4}[source]
I think the fan noise you experience is not Apple's fault, but Intel's fault.
replies(1): >>25071154 #
22. torbital ◴[] No.25066717[source]
But now they are locking down their software hard. So there is really no free lunch.
replies(2): >>25066753 #>>25066822 #
23. the_lucifer ◴[] No.25066753[source]
Someone here posted when the M1 powered Macs launched that they're essentially ASICs now and that made me do a double take. Since you cannot change the operating systems anymore, aren't these essentially ASICs?
replies(1): >>25066992 #
24. jariel ◴[] No.25066772[source]
This is salient, and almost upsetting frankly as I (and others?) have been looking for a 'way off' the platform after years of grievance. This is might just be good enough to keep their core platform value in place. It's a shrewd move in their part, it's been a while since we've seen this level of core innovation on their non-iOS offers.
replies(3): >>25066863 #>>25067355 #>>25067560 #
25. avalys ◴[] No.25066822[source]
In what respect is Apple “locking down their software hard” with respect to these new Macs?
replies(4): >>25067018 #>>25067027 #>>25067452 #>>25069725 #
26. silly-silly ◴[] No.25066833{3}[source]
Start adobe premier, any significantly complex scene and render it. The system will throttle down to 1.2ghz, sometimes even 1 ghz .

I'm testing on is the 2019 core i9 2.3 ghz. Apple "Genius" bar tells me this is expected, no replacement in warranty. Maybe the 2020 is better, but after getting this as an answer I wont be buying another apple macbook.

The prior 2018 Core i9 Apple MacBook Pro had previous issues with insufficient power supply to the CPU core (this might be fixed, but I can't even sustain it due to thermal throttling, so I can't really tell).

27. acomjean ◴[] No.25066863[source]
It’s a huge problem for open computing to be sure.

This isn’t risk free, issues having to do with supply and process and frankly geopolitics can cause problems. But it looks like they’re off to a good start.

28. fiddlerwoaroof ◴[] No.25066880{3}[source]
I’ve done this exercise a couple times: pick a random Apple laptop configuration and find an equivalent from another vendor: they’re frequently so close price-wise it doesn’t matter (depends on what you want, but I’m willing to pay $200 or so for macOS and not having to deal with Linux power management.)
replies(1): >>25069942 #
29. asimpletune ◴[] No.25066992{3}[source]
Not even close. An ASIC is a program on a chip. These are still general purpose, programmable computing machines.
replies(1): >>25067658 #
30. sooheon ◴[] No.25067018{3}[source]
As a long time Mac user, the last few OS updated caused me to have to good some permission error or other just trying to run simple tools like emacs. They know they can't control the web and web apps, but they are making moves to lock down apps that run on their OS.
31. Polylactic_acid ◴[] No.25067027{3}[source]
Its likely the end of the days of installing an alternative OS. Bootcamp support has been dropped and Linux support will likely not exist.

And then there is the OS which is getting more and more locked down so that you can not run unsigned software without increasingly difficult workarounds.

On one hand, alternative OS support on macbooks has gotten worse and worse over the last few years but it is sad to see the final nail in the coffin.

32. totalZero ◴[] No.25067074[source]
Not quite. You can't synthetically compare one chip to another and draw organic conclusions. Apple computers with M1 don't support the software that I use. That's why I bought a fully spec'd Intel MBP13 today. All this talk about battery life and benchmarks gets flipped on its head when I can't use the product in the real world.
replies(1): >>25067232 #
33. manojlds ◴[] No.25067232[source]
Isn't there Rosetta that they talked about?
replies(2): >>25067655 #>>25069704 #
34. kccqzy ◴[] No.25067331[source]
I've personally found Geekbench results representative of real time workloads. So do more traditional benchmarks like SPECint.
replies(2): >>25067664 #>>25068327 #
35. hackerfromthefu ◴[] No.25067355[source]
I'd expect their performance features, such as the wide execution and L2 etc, to be copied by amd/intel within a couple of years.
36. josteink ◴[] No.25067452{3}[source]
> In what respect is Apple “locking down their software hard” with respect to these new Macs?

Locked bootloader only booting stuff signed by Apple.

So these CPUs can only be used to run MacOS, no Linux or other alternate open platforms.

replies(1): >>25068474 #
37. bsaul ◴[] No.25067560[source]
same feeling here. It seems that we'll soon have to choose between freedom on one side vs performance ( inaddition to good UX, which is today's tradeoff).

i was hoping for a good ios opensource replacement, i guess now i'll soon have to wait for a good laptop competitor as well.

I feel like i'm stuck in a jail made out of gold.

38. AntiImperialist ◴[] No.25067563[source]
They have 5nm chips because they essentially outbid everyone else from doing so at this time and bought all of the production capacity from TSMC for themselves.

They could do that because they've been selling overpriced products.

39. 0x008 ◴[] No.25067655{3}[source]
And with rosetta it is certainly slower than running it native on x86
replies(1): >>25068038 #
40. qeternity ◴[] No.25067658{4}[source]
An FPGA is a program on a chip. An ASIC is literally just a chip. Calling the M1 an ASIC is insulting to the Apple hardware team.
replies(2): >>25068263 #>>25068277 #
41. michaelmrose ◴[] No.25067664{3}[source]
Fantastic since many pieces of kit have been released to the public wherein both geekbench and more traditional measurements of actual performance exist can you point out some instance where real benchmarks are well correlated with differences in geekbench score?
42. manojlds ◴[] No.25068038{4}[source]
Maybe, maybe not?

https://twitter.com/Catfish_Man/status/1326238434235568128

43. patrec ◴[] No.25068229[source]
> Now Apple computers [...]

Apple no longer sells computers. You can rent some shiny gizmo from them to run software of their choosing provided by people they deign to allow on their platform and in a manner they approve of¹. It's not really yours anymore.

¹ "But you can still do X". Well, and last year you could still do W, and the year before that V.

44. guax ◴[] No.25068263{5}[source]
Yes, M1 is not just a chip, is a magical piece of the rib of steve jobs himself. I agree though its not an ASIC even though it have ASIC things inside of it, just like any other modern processor.

It is much more complete SoC then other procs which makes its performance even more impressive if this indications hold up, I am still very skeptical, nothing comes for free and the real world is a b

45. zaro ◴[] No.25068277{5}[source]
ASIC stands for application specific, and since thei IC is application specific - used only in certain hardware along certain software, it's kind of correct to call it ASIC.
replies(1): >>25068768 #
46. sirn ◴[] No.25068327{3}[source]
Anandtech has done SPECint on A14 and result is consistent with what we've seen from M1 on Geekbench. It's not the same CPU, but they share the same Firestorm/Icestorm cores.

https://www.anandtech.com/show/16226/apple-silicon-m1-a14-de...

47. xenadu02 ◴[] No.25068474{4}[source]
Where are you people getting this from? This is not accurate. You can in fact downgrade security from recovery OS.

ARM doesn't have a generic platform like PC but I'm sure someone will figure out how the device tree works if they haven't already.

replies(1): >>25069899 #
48. rusticpenn ◴[] No.25068768{6}[source]
I am unsure, since the application here is general purpose computing.
49. soneil ◴[] No.25068824[source]
I think it's more that we want information today and we don't have actual user testing today. And we're all guilty of this - why else did we click when the machines are still pre-order.
50. totalZero ◴[] No.25069704{3}[source]
It's an emulator, not a panacea.

"Rosetta is meant to ease the transition to Apple silicon, giving you time to create a universal binary for your app. It is not a substitute for creating a native version of your app.”

https://developer.apple.com/documentation/apple_silicon/abou...

replies(1): >>25070327 #
51. delfinom ◴[] No.25069725{3}[source]
Apple is moving to basically ban non-Apple store installs of Mac software. It's been in progress the last few years but they are on the final stages of turning Macs into iPhones.

Gotta take their 30% cut of everyone's revenue.

replies(2): >>25069972 #>>25072842 #
52. cesarb ◴[] No.25069899{5}[source]
ARM does have a generic platform like the PC, it's the BSA (https://developer.arm.com/documentation/den0094/a).
replies(1): >>25120638 #
53. ProAm ◴[] No.25069942{4}[source]
There was a period of time where the laptops were not updated for almost 5 years and the price didnt change. No one remembers the past
54. ralfd ◴[] No.25069972{4}[source]
In the release event they featured prominently Cable Sasser from Panic and an App which is not in the App Store.

https://twitter.com/cabel/status/1326271980081876992

55. manojlds ◴[] No.25070327{4}[source]
It's a stop gap yes, but I am replying to someone that said their apps will not run on these new machines.

And check my reply to a sibling comment where people are finding cases where Rosetta is faster than native.

replies(1): >>25078505 #
56. faitswulff ◴[] No.25070415[source]
Judging by upward trajectory of their processors’ performance (https://images.anandtech.com/doci/16226/perf-trajectory.png) and their dominance in mobile, it seems like only a matter of time.
57. wishysgb ◴[] No.25071110[source]
wait for it. this all sounds too good to be true. I would love it if it was and I will be the first to get a Mac but it doesn't sound true.
58. wishysgb ◴[] No.25071154{5}[source]
well, my very powerful laptop barely every gets any noise. Most of the time it is silent until I am compiling something. It is of course apples fault as their cooling is suboptimum
replies(1): >>25073397 #
59. lasagnaphil ◴[] No.25072842{4}[source]
Probably not, since there are many crucial pro-level applications related to image editing, modelling and animation, video and audio production that realistically wouldn't be on the App Store (well, apart from Apple's own products).
60. kube-system ◴[] No.25073397{6}[source]
Is Apple's cooling suboptimum or is Intel's TDP suboptimum?

"Make it thicker and heavier" is apparently not the answer that Apple was looking for from Intel.

Thin and light Wintel PCs are known for having a lot of thermal issues too.

61. totalZero ◴[] No.25078505{5}[source]
Actually, you're replying to me. Check the usernames.

You can't run Windows on these things, and Rosetta 2 doesn't fully support kexts, VMs, or certain instruction sets. It's a translator and it's going to be imperfect in practice. That's why it's not intended to supplant development with native instructions.

Your other comment is a tweet regarding one function that is speculatively faster, but tells me nothing about real-world performance -- nor whether the tools I use for my business are going to be supported by Apple Silicon in the next few months.

62. xenadu02 ◴[] No.25120638{6}[source]
LOL, I should have known someone would try to "well akshually" me.

The BSA/SBS is relatively new as far as I'm aware. The server version was released in 2014, the same year as the iPhone 6 which was already using Apple SOCs.

I don't know when the client version was released but fairly recently AFAIK. I don't know of any systems shipping based on it.

Most ARM systems are using device trees and their own custom slate of devices.

So I should amend my comment I suppose: no one is using any kind of "Standard ARM PC" definition in any quantity, and I'm not sure we should bring over UEFI or ACPI when device trees have been working well so far.

Nevertheless as I noted I'm sure enterprising hackers will figure out how to do it. If you downgrade security the SEP will sign whatever "kernel blob" you like and the system will load and jump to it at boot. Technically that isn't even required - a kext could pause all CPUs, set the correct registers, and jump to a completely different OS if you were really determined.