ARM MacBook Pro I don’t want an Overpriced Netbook.

[Jorbanead]

Let’s keep it simple: Apple would never release a new ARM MacBook that was worse than the year before. Their entire MO is “the new iPhone 11 is 30% faster and 50% more efficient than the iPhone X”. They always compare the latest and greatest model with last years, and show how much better this new one is.

Plus, Apple has a lot to prove. The whole tech world is waiting to see what happens with this transition, and Apple knows if they couldn’t come out with anything other than a great product, journalists would have a field day with “Apple is doomed!”. There’s no chance in my mind that all AS macs are going to be AT LEAST just as good as the current offerings - and more likely better than the current intel macs.

 

[filu_]

You are right in principle, but what is this "better" about? Maybe 30% longer battery life and 15% less weight (after throwing out the cooling tube)?

 

[UpDown]

Let’s keep it simple: Apple would never release a new ARM MacBook that was worse than the year before. Their entire MO is “the new iPhone 11 is 30% faster and 50% more efficient than the iPhone X”. They always compare the latest and greatest model with last years, and show how much better this new one is.

Plus, Apple has a lot to prove. The whole tech world is waiting to see what happens with this transition, and Apple knows if they couldn’t come out with anything other than a great product, journalists would have a field day with “Apple is doomed!”. There’s no chance in my mind that all AS macs are going to be AT LEAST just as good as the current offerings - and more likely better than the current intel macs.

You've forgotten about the 2014 Mac Mini, which was half the performance of the 2012 Mac Mini, and it stayed that way until they introduced the 4th gen in 2018.

 

[raknor]

So point #3 is the only drawback of larger page size. However, I know for sure my threads will read almost exactly the same sized chunks of data, and will free up that chunk when they're done, so there is no waste.

These statements make it abundantly clear who doesn't quite get it. Read up a bit more then may be you will get the point I am making.

The boost from Hyper-Threading just highlights the inefficiency of Geekbench's multi-core test. It's not a highlight of Apple's approach.

Not remotely. Apple's approach clearly shows the Performance cores in the A12Z handily outperform the Core in the i9 when SMT is not used in both Single core and Multicore.

I'm not "going to consider" anything. The A12Z is an 8-core CPU and it is a fact that all 8 cores can be active at the same time. They are physical cores.

There is nothing for you to consider, 4 of those cores are not at the same performance level. At least that much must be clear to you by now.

Hyper-Threading creates virtual cores, but the exact count of physical processor cores is still 8 no matter how you want to slice it.

Sure. If you are impressed by the 30% boost that hyper threading lends and you shouldn't be disregarding the 25% boost the low performance cores in the A12z either.

I am referring to this statement

Without the 25% performance boost, the A12Z would only score about 3640. So I think it's far less impressive without the LITTLE cores.

The Core i9-9980HK in my MBP 16" with 4 cores and no hyper threading gets a similarly less impressive 3700. BTW this is with Turbo enabled, we know without turbo and at 2.4 GHz it gets an even less impressive 2365.

It is sad that the Core i9 in my MacBook Pro 16" hits 68 watts peaks for the cores during geekbench multithread and is not that much faster than a 10W Apple chip. An A12z with 8 performance cores would obliterate the i9 while consuming a hell of a lot less power.

The results speak for themselves anyway you slice it and that's why people are eagerly waiting for the Apple Silicon based macs.

 

[johnsc3]

This is the same complaint computer science neophytes said in 2015. Apple needed to do this years ago. People are just delusional because they think Intel is a good chip-designer when they’re really in 3rd place. Behind AMD in 2nd, which is behind Apple’s microarchitecture design abilities in 1st. Apple’s chips were comparable architectures to Intel’s back with the A6 or A7 chips.
Apple’s chips surpassed Intel’s in performance-per-watt and maximum capable performance years ago. The only thing I worry about is them skimping on how many cores they put in it. I’m pretty sure they could put 12 high performance cores and 4 high efficiency cores in it and with that it’d be pushing beyond the capabilities of their iMacs. But reports are they’re only doing 8 high performance and 4 high efficiency cores in the first chip for the 12” macbook. I’d like them to have a performance option.

Regardless, if they clock their macbook chips with 3+ghz turbo they’ll obliterate intel’s chips. And they can do that because a laptop has much better heat dissipation and battery capacity than an iPhone or iPad.

the most exciting possibilities are when the do a multi-chip package for desktops with fans. They’ll have the multicore capabilities of massively powerful chips in a TDP using half the watts and making half the heat. Plus, it’ll be cheaper to make. Probably would use pci-express 4.0 or 5.0 as the chip-to-chip interconnect.

It is as if some folks don’t remember or realize how groundbreaking the debut of the A7,the first 64 bit mobile processor. If chip makers And others didn’t take Apple seriously with the A4, the A7 certainly put everyone on notice.

 

[dogslobber]

Unless these ARM chips are 2-3x faster than 10th gen Intel equivalent then there is no value proposition. People bought into Intel Macs due to the safety net of being able to run Windows if need be. That's one of the reasons why PowerPC only sold a couple of million systems. Nobody wanted to be stranded by a platform with no peers. They same issue exists with ARM where Apple is spreading into chips used to run washing machines. Unless these chips have a wow factor off the bat then they will be viewed with the woe factor. A general rule of thumb is expecting a 2-3x performance increase to allow early adopters to step onto the ledge. There is safety in numbers which is why it's likely that a weaker ARM chip will leave Intel Macs INCREASING in price. Stay tuned.

 

[dmccloud]

Nobody wanted to be stranded by a platform with no peers. They same issue exists with ARM where Apple is spreading into chips used to run washing machines.

You conveniently forgot Microsoft's Surface Pro X, Samsung Galaxy Book S, Lenovo Mixx 630, Lenovo C630, and the recent reports that Samsung is now bringing their ARM-based Exynos processors to the PC side of things. There is a push on multiple fronts to move away from x86 and Intel on both the Windows and MacOS side, and many Linux distros also support ARM-based systems. To equate Apple Silicon to "chips used to run washing machines" is ridiculous and shows just how little understanding you have regarding the differences between the chips Apple, Samsung, and Qualcomm are producing for smartphones and computers and the chips Maytag puts into their cheap dishwashers.

 

[leman]

Unless these ARM chips are 2-3x faster than 10th gen Intel equivalent then there is no value proposition.

I’ve been buying a new MBP every two years for meager 10-20% performance increase. Because it’s worth it for me. Even if the Apple Macs end up only 30% faster it will be more than worth it for many users. You are overestimating the number of people who run Windows on Mac.

 

[ericwn]

Unless these ARM chips are 2-3x faster than 10th gen Intel equivalent then there is no value proposition. People bought into Intel Macs due to the safety net of being able to run Windows if need be. That's one of the reasons why PowerPC only sold a couple of million systems. Nobody wanted to be stranded by a platform with no peers. They same issue exists with ARM where Apple is spreading into chips used to run washing machines. Unless these chips have a wow factor off the bat then they will be viewed with the woe factor. A general rule of thumb is expecting a 2-3x performance increase to allow early adopters to step onto the ledge. There is safety in numbers which is why it's likely that a weaker ARM chip will leave Intel Macs INCREASING in price. Stay tuned.

This is not 2005. Apple has changed and its customer base has as well. I don’t believe for a moment that this change will impact them negatively for anything more than the immediate switch moment for a a small subset of users.

 

[raknor]

Unless these ARM chips are 2-3x faster than 10th gen Intel equivalent then there is no value proposition.

Why? Is Intel's 11th gen chip 2x-3x faster than the 10th gen? Was Intel's 10th gen 2x-3x faster than the 9th?

This straw man of creating an extremely high bar to meet is very dubious and blatantly obvious.

People bought into Intel Macs due to the safety net of being able to run Windows if need be.

No. An exceedingly vast majority of Mac users don't use Windows on their Macs. I have VMWare Fusion and haven't used Windows in years. My VMWare ESXi home lab is running Linux instances too.

That's one of the reasons why PowerPC only sold a couple of million systems. Nobody wanted to be stranded by a platform with no peers.

There are over 3 billion ARM devices in peoples hands and lives (Android, iOS, IoT).. there are 1.5 Billion PCs running windows on x86. No peers, my ass. When PowerPC transitioned to x86, x86 was the dominant CPU architecture, period. ARM is now the dominant arch.

 

[dogslobber]

I’ve been buying a new MBP every two years for meager 10-20% performance increase. Because it’s worth it for me. Even if the Apple Macs end up only 30% faster it will be more than worth it for many users. You are overestimating the number of people who run Windows on Mac.

People won't buy ARM Macs that are hobbled to being only 30% faster than the current generation of Intel Macs. There's no value proposition to be had. You're getting confused between the perception of running Microsoft and those actually doing it. Those are totally different numbers.

 

[throAU]

Not in single-threaded tasks.

They'd take the opportunity to make more money though!

Dunno about that. Server chips need to be larger and larger is more difficult to manufacture and would detract from their mobile device processor manufacturing which is already likely contained by TSMC capacity. Apple has found a niche and are making hay with it.

 

[throAU]

People won't buy ARM Macs that are hobbled to being only 30% faster than the current generation of Intel Macs. There's no value proposition to be had. You're getting confused between the perception of running Microsoft and those actually doing it. Those are totally different numbers.

I guarantee you they will if they are fan-less, even if they are only the same or marginally better performance, with less heat and better battery life.

Which they will be.

Portables are the vast majority of Apple's market, and tasks that need to scale up to the bigger hardware (read: Mac Pro) are generally multi-threading friendly. Which means Apple can simply scale up with raising the core count and/or task-specific programmable ASICs/FPGAs like in the afterburner card.

 

[throAU]

Because the Air is limited to <10W of thermal dissipation. It's not able to properly allow max sustained performance even with the iPad Pro's A12X.

Nah, it's not. The total system power draw on the MBA is up to 30 watts or so. The processor alone in the MBA is rated at 12 watts TDP (the i7 model, which is outperformed by the iPad Pro) and regularly sees more than that during boost.

The iPad Pro runs in 12-15 watts.

Fun fact: I thought ARM Macs are not as bad as well, until I tried to simulate the behavior of virtual memory and running multiple different processes on my iPad Pro. It's not pretty. This is from real life experience, not from "data".

Fun fact: you're not running on the ARM Mac hardware, you're running on something multiple years old and designed for single tasking with limited RAM. Your actual real life experience doing that has zero relevance to future unreleased ARM hardware.

 

[Emanuel Rodriguez]

The new MacBooks will blow away Intel Macs.

Performance for Safari and other first tier apps will be fantastic.

Other apps, that will rely on Rosetta 2 for 1-2 years before they’re updated, will have performance on or or slightly better than current Macs.

If you need bootcamp or gaming, you may buy the Intel Macs that are still on sale

Yes, the current evidence suggests that all of this is true. I do hope that some sort of Bootcamp solution ends up being possible with ARM Macs. It certainly is technologically possible, it's just a question of will.

 

[Maximara]

People won't buy ARM Macs that are hobbled to being only 30% faster than the current generation of Intel Macs. There's no value proposition to be had. You're getting confused between the perception of running Microsoft and those actually doing it. Those are totally different numbers.

There is more then just "30% faster". Lower energy use comes to mind which means you don't have the performance drop seen with Intel Macs because Intel couldn't produce the chips Apple had designed its Macs around and so heat throttling out the wazoo.

 

[throAU]

Yes, the current evidence suggests that all of this is true. I do hope that some sort of Bootcamp solution ends up being possible with ARM Macs. It certainly is technologically possible, it's just a question of will.

It won't.

Boot-camp is dead with the ARM transition, unfortunately if you need it you're going to be limited to a cloud desktop (for business apps) or a gamin console/PC for stuff like that unless you're willing to live within the Apple software library.

It's unfortunate, but Apple isn't going to emulate X86 at a hardware level required to enable boot-camp.

 

[slooksterPSV]

It won't.

Boot-camp is dead with the ARM transition, unfortunately if you need it you're going to be limited to a cloud desktop (for business apps) or a gamin console/PC for stuff like that unless you're willing to live within the Apple software library.

It's unfortunate, but Apple isn't going to emulate X86 at a hardware level required to enable boot-camp.

I wonder what the licensing costs for that would be? I mean the x86 instruction set has to be licensed to be used I haven’t looked at ARM to see if it’s similar but I know Intel owns the rights to x86 so Apple would have to license it to run x86 at a hardware level where Rosetta is a translation. You know what.... maybe they are already licensing x86 for Rosetta 2.

 

[throAU]

It's not so much licensing cost that is going to prevent it, it's power consumption, heat, die space, etc. for hardware that doesn't encourage migration.

Apple don't want you running x86 software any more, and while the hardware is out there to do so they will be carrying the maintenance of it all.

The number of customers they will lose by abandoning it is not significant enough to warrant the compromises they'd be making to enable it.

Rosetta install-time translation is good enough to cover the 99% case.

 

[johngwheeler]

People won't buy ARM Macs that are hobbled to being only 30% faster than the current generation of Intel Macs. There's no value proposition to be had. You're getting confused between the perception of running Microsoft and those actually doing it. Those are totally different numbers.

I would hardly say a 30% improvement in a CPU these days is "hobbled" - it's a pretty healthy generational improvement that we haven't seen in CPUs for some time. Would I like ARM to be 2-3x faster than Intel? Of course! But I don't think most people buying new Macs are expecting it or demanding it. I expect that the vast majority of buyers, a simple "here's the new Mac - it's 30% faster and the battery lasts 15 hours" would be enough...

I'm curious as to what "the perception of running Microsoft" means? I wouldn't think it is a matter of perception - you either are, or you aren't...

I've seen statistics that claim only 2% of Mac users run Bootcamp. No idea if it is correct today, but with the spread of platform-independent web-apps (Office 365 etc.), I can't imagine that this is a growth area. I moved from Bootcamp to running VMWare and then to cloud instances or separate Windows machines when I need to use Windows.

I really don't think new Mac buyers are going to be really bothered by this unless they are dependent on specific Windows apps and can only use one local machine. Gamers will opt for a dedicated PC or console given the choice, and there are MacOS alternatives to most Windows apps. I have some Windows-only astronomical control software (for imaging) and it was much less hassle just to buy another dedicated Windows computer (Intel NUC) that to try to get everything working on my Mac.

 

[throAU]

I would hardly say a 30% improvement in a CPU these days is "hobbled" - it's a pretty healthy generational improvement that we haven't seen in CPUs for some time. Would I like ARM to be 2-3x faster than Intel? Of course! But I don't think most people buying new Macs are expecting it or demanding it. I expect that the vast majority of buyers, a simple "here's the new Mac - it's 30% faster and the battery lasts 15 hours" would be enough...

Yeah if people didn't buy without a 30% speed improvement we'd all still be on sandy bridge, if you were hold intel to the same standard.

Even same speed and 15 hour battery with no fan would be a MASSIVE selling point. Fans are loud. What's the biggest MacBook complaint on this forum? Jet turbine fan noise.

 

[richinaus]

Yeah if people didn't buy without a 30% speed improvement we'd all still be on sandy bridge, if you were hold intel to the same standard.

Even same speed and 15 hour battery with no fan would be a MASSIVE selling point. Fans are loud. What's the biggest MacBook complaint on this forum? Jet turbine fan noise.

Fan noise, heat and battery life are the top 3 complaints with the MBP. I dont recall reading anyone disappointed with the speed.
If it was 10% slower but nailed the 3 complaints, I would still get it [plus running my apps natively, not Rosetta].

There is one thing I have learned on Macrumors though - don’t believe anyones opinions, test it yourself. Rosetta especially is one thing I will have to test for myself over believing any comments on here, like ‘it runs very smoothly’ when in fact it is a P.O.S. One persons judgement of speed / fan noise / heat etc can vary a lot to another

 

[leman]

I dont recall reading anyone disappointed with the speed.

I am disappointed with speed. In single-threaded performance, there have been barely any advances in the last couple of years.

 

[dmccloud]

It won't.

Boot-camp is dead with the ARM transition, unfortunately if you need it you're going to be limited to a cloud desktop (for business apps) or a gamin console/PC for stuff like that unless you're willing to live within the Apple software library.

It's unfortunate, but Apple isn't going to emulate X86 at a hardware level required to enable boot-camp.

Why would BootCamp be dead? Microsoft already has Windows on ARM, and there have been multiple reports of discussions between Apple and Microsoft regarding possible licensing to allow WoA to run on Apple Silicon Macs. Since Microsoft already has an ARM-based version of Windows that is also being run on the Samsung Galaxy Note S and several Lenovo machines, the porting and driver work is largely done. All that would be needed for BootCamp are some Apple-specific drivers and a licensing arrangement between the two companies. This also means no need for x86 emulation or licensing arrangement to allow for running or emulating x86 code.

 

[dmccloud]

Fun fact: I thought ARM Macs are not as bad as well, until I tried to simulate the behavior of virtual memory and running multiple different processes on my iPad Pro. It's not pretty. This is from real life experience, not from "data".

And then going through the "data" just shows that Apple has come a long way, but they are not "ready" yet to replace the absolute performance offered in the 16" MacBook Pro, the iMac, and the Mac Pro. That's what I've been saying all along.

So you tried to "simulate" virtual memory and running multiple processes on your iPad, and that is somehow representative of what Apple Silicon can do under MacOS? Your "data" is flawed, misleading, and completely irrelevant because you ran it on an older A-series processor under iOS instead of actual Apple Silicon running MacOS. Protip: the DTK is far more powerful that you realize, and that's on the older A12 series processor. You drop a processor designed for MacOS into that machine and I'd wager it will absolutely scream, and will definitely perform better than whatever you were trying to accomplish with your non-representative testing "methodology".