Introducing the next generation of Mac - A new MacBook Air, 13-inch MacBook Pro, and Mac mini powered by M1, Apple’s chip designed specifically for the Mac
This data is scraped automatically and may be incorrect.
- Introducing the next generation of Mac
- Word count
- 2687 words
what keeps the rest of the computer world to move to ARM longterm?
Windows on arm already has a comparability layer for x86_32 to arm. They are working on x86_64, but I’m not sure when they are releasing it.
Also, Apple has the ability to throw a processor that is strictly better in many ways than anything intel can create (at least for laptops. We will see what Apple can do for desktops). Even if Rosetta 2 isn’t very performant, the m1 can brute force through to good performance. I don’t think Qualcomm or Samsung have anything that can come close.
The base model MacBook Air gets a slightly cut down GPU (7-core vs 8-core) and is almost certainly a lower binned part with lower clocks (for the fanless design). Nothing unexpected or surprising but it would nice if Apple said how much lower the clocks are.
The most interesting thing here is the “unified memory” design. It certainly looks like HBM, which is frankly a big deal.
I don't know if it has been confirmed that Apple is using HBM instead of more standard DDR4 that just happens to be close to the CPU, so this is just speculation on my part.
This is about what I expected, right down to the name of the chip—M1—which replaces the previously used M prefix for the motion co-processors in the A-series chips. Eventually those were integrated into the die of the A-series, but you have to wonder whether Apple realised they'd made a slight boo-boo right around the time they internally decided to head towards ARM-based Macs by using the
Mprefix on something so dull as a motion chip in a phone!
I'm shocked about the lack of product segmentation here, both between individual product lines and the series of notebooks themselves. Take the MacBook Air for example: there's some binning going on in that lineup with the 7 core versus 8 core GPU, but it doesn't seem like enough of a differentiator to force people to upgrade? That's a lot of binned chips. Then there's the MacBook Pro 13", with its "integrated" graphics and the exact same chip as the MacBook Air—except in this instance, it's actively-cooled. You just have to "know that", I guess? And infer that you're going to get better sustained performance from the actively-cooled model.
My question is how is that any more obvious that just providing maximum clock speeds? Apple's rapidly encroaching in on the territory where they're losing the ability to differentiate their products easily because they're so determined to obfuscate technical specifications. They had this occur with the iPhone's this year, for example. Because they refuse to discuss RAM in their phones, you wouldn't know that the iPhone 12 has 4GB, and the iPhone 12 Pro has 6GB. The entirety of the differentiation came down to, essentially, camera features and materials.
I both understand it and I don't. On one hand, yes, sure, you don't need to know how your computer works, it's there to do work—but it's not entirely a hermetically sealed abstraction. We want that information to be able to make informed decisions around our purchases. It's like Tesla removing the kWh ratings from their vehicles and dropping back to "standard range" and "long range". Frustrating and irritating. There isn't really an issue for product differentiation around the Mac mini here, because the Mac mini is its own class of consumer desktop, and its own unique form factor. That isn't true for the MacBook, which are all—fundamentally—laptops at the end of the day.
I don't understand the constant gripes about not updating the design or around repairability here (the latter is especially true given Apple's stated goals of using recycled materials and creating an end-to-end recycling loop to accept old devices as a way of making new ones). There was clearly a conscious choice to ensure people feel at home with their existing designs, and knowing that these new devices with new internal architectures will just work, although one can't help but lament the continued existence of that fucking Touch Bar.
If you don't know how it works, I think you're just supposed to assume that a Pro goes faster than an Air.
The differentiator between the Air and the Pro (besides the binning) is the presence of active cooling. The Pro will be much more suited for consistent workloads like video rendering, machine learning, and video games. The Air might be able to hold high clock speeds for a minute or two but then it will throttle back down to base clocks as the thermal solution won't be able to keep up.
If your use cases are reliant on quick bursts rather than sustained workloads, then the Air definitely seems like the better value.
Design - a little disappointed that it's literally the exact same chassis, but to be honest there wasn't any reason to expect this year to be different. Still has the largeish bezels, a 720p webcam, and no touch screen. I think next year's the year for an actual redesign.
Performance - who knows, but I believe them on battery life at least. 20 hours for video watching is pretty amazing. I think that Baldur's Gate footage is pretty promising too; the minimum graphics spec for that is a GTX 780, recommended is a 1060. If that Mac Mini is between 780 and 1060 levels performance... that's actually a good deal. It would genuinely be among the high end for 13'' laptops. edit: I realized this was confusing; the Mac mini, being $700, would be a pretty solid deal for that performance; the macbook pro would have fairly high tier graphical capabilities for 13''.
Obviously performance from these Apple presentations should be taken with several large grains of salt, so I'm excited for reviews to come out next week.
I'm not particularly worried, these are fundamentally beefed up iPhone chips, evidently throttled enough it's nowhere near hot enough to cause such damage.
Well, yeah, that's why they only showed BG3 on the Mini and Pro, where it does have a fan. Likely the main performance difference is that the Pro won't have any throttle at load.
Also, the comparison to the 780's power isn't exactly fair. That's a GPU from 2013. If you just compare even Intel's latest integrated graphics to its discrete equivalent, there are extreme gains in efficiency. And according to apple, they're 2x more efficient than something.
We'll have to see once it comes out, but maybe it's something akin to turboboost. Although I hate turbo boost because of the immense heat it generates -- I actually turn it off on all my laptops (MBP 2015, 2018) so they run cooler since the boost in speed isn't really noticeable for my workloads.
That's one aspect that I'm not too incredibly worried about; with that kind of energy efficiency thermals don't tend to be much of a problem.
Anand tech coverage, much more deep into the technicals than other articles.
In particular, it's truly insane that the iPhone's cpu is roughly equivalent in performance already to the top of the line Intel and AMD CPUs. Might see some truly spectacular numbers when the hardware comes out.
It's not. Look at this graph from the article, for instance
That ain't no mobile chip. That's barely losing to the Ryzen 9 5950X, and beating the i9 10900k. Desktop class CPUs with 10x the power usage.
The 5W A14 seriously competes with a 50W Ryzen CPU. Hence why it's mind boggling.
There are some leaked benchmarks now of the M1, which are mostly inline with what people expected (slightly better single core perf than the A14, which note is incredibly high, and multicore that is scaled up from the A14).
But they would mean that the new Macbook Air (which was specifically the leaked benchmark) is strictly stronger in CPU in both single core and multicore than the i9 on the 16 inch MBP. Which would suddenly make the Air from a meh price/performance to probably the best price/performance laptop in the market.
Are these leaked benchmarks supposed to be reliable? I'm a bit skeptical at the moment because I noticed that Apple is also currently selling the equivalent Intel based MBP at a higher price. If I were Apple, I'd be trying to keep the price high if the performance is really that great.
That being said, since the majority of software at launch is still going to be written for x86, I'm sure there's going to be a fair amount of performance penalty to negate these numbers, and perhaps that lack of confidence is why Apple doesn't want to keep the price up. Beside that, I'm sure they're making plenty of profit simply because they aren't paying the Intel tax.
Not particularly, but they are about what you'd expect. The iPhone chip already performs better than the i9 10900k (yes, the desktop one) and only slight behind the Ryzen r9 5950x (yes, the 50w desktop one) in single core performance, and it has about 3800 multicore score with 2 high performance cores.
So the single core performance is slightly higher (but which would put it as the highest single core performance on any consumer CPU), and the multicore is about 2x what the iPhone chip has (since the M1 has 4 high performance cores vs 2 on the A14).
As to price, I mean Apple needs to prove to consumers why this is a worthwhile thing to do.
Emulated benchmarks have come out, testing the M1s performance after going through Rosetta 2.
They show a predictable drop in performance, but surprisingly, they still outperform every Intel MacBook. Even when emulating x86-64. Good sign for the prospect of running all the software that hasn't been recompiled for ARM.
$700 for the Mini is fairly cheap for a desktop Mac, but Apple's tech specs aren't very useful. I will wait for an independent review.
You know what, I've always bitched that Apple's Mac Mini pricing was always way higher than when it first came out, but if you look at inflation, this one is only about $50 more and comes with what's supposed to be a high-end processor.
If you believe the marketing (and I for one would be highly cautious as these are really first-generation devices), the pricing on the laptops are really not bad either. $999 is pretty choice pricing for the Air model.
I seriously considered buying one this morning, but it's an extra £200 for 16GB RAM, which killed it for me. The almost-decade-old Mac Mini it'd be replacing already has 16GB RAM and 256GB SSD, and is honestly still a pretty great machine, it just can't drive modern displays at full resolution/refresh rate - £700 for a solid upgrade, better thermals, better connectivity, plus the nerd fun of playing around with the new architecture was pretty close to a sale for me. Pushing it up to £900 or having to take a step back on memory (again, nearly 10 year old machine) was enough for me to hold off for the moment.
Yeah they talked me off the ledge with building a PC instead with the Mini announcement, but I think I’d still rather wait a year or two before making the switch. Hopefully they’ll have a much better chip and eGPU support in the next iteration.
Which is a shame though. I was looking forward to upgrading this year as my current iMac is getting a bit long in the tooth.
Very interesting...I have high hopes for Apple Silicon. It looks interesting and I'm excited to see how they deal with some of the higher end computers (especially what they'll do with 3rd party graphics cards). I like the ability to run iOS apps. I think that'll be a big attractor for people who own an iPhone but use a Windows PC.
I'm also disappointed they didn't announce cellular support or touch screen support. Hopefully we'll get a major design refresh soon that adds those features.
I do think it's a good thing they kept the external appearance the same, to help ease the transition.
But the thing that jumps out at me is that two of the biggest issues haven't been solved by this transition:
It doesn't matter that a computer has an SSD that's two times faster if it's full within a week and you're constantly relying on external drives for storage.
8 GB of RAM and 256 GB of storage on a $1,300 laptop is shameful. It's fine for the most basic of uses but someone who is buying a MacBook Pro probably isn't looking for something "basic". And RAM also maxes out at 16 GB, which seems like a weird step backward for a laptop who's previous version supported up to 64 GB of RAM.
I feel like it goes in cycles with Apple where they raise their minimum specs, then they wait way too long to update the specs again. During that time, resentment and annoyance grows until a few years later when they bump up their specs. It feels like we're at the peak of that wave where they are way behind everyone else, sort of like we were for the last few years with the 16 GB iPhones. Until earlier this year, Apple was still selling laptops with 128 GB of storage.
For RAM, the X factor is that it's using shared memory, and possibly HBM for that matter. It looks like it's on the SoC itself. For normal laptops, increasing RAM is as simple as sticking more of it into the motherboard. That probably isn't low latency enough to work as both ram and vram.
Hence why you can't go >16 GB no matter how much money you throw at the configurator.
I get that - but the technical limitations aren't the problem of the end user. Apple has to know that eventually they'll need to make laptops with more than 16 GB of RAM. Unified memory does change the equation somewhat, but like you said, it will depend on type/location.
It's the same problem they created for themselves when they used LPDDR3 RAM and had a 16 GB RAM limit in the previous generation of laptops.
I really hope this spurs a greater movement towards ARM on laptops. I want to buy an ARM linux laptop, but the only ones out there right now are very budget focused. The new macbook is very close to what I'd want, but my ideal laptop would be more user serviceable and have ports that aren't USB type C.
It practically already has. It's going to create market pressure, because these ARM CPUs have a killer feature that is going to make other manufacturers want them: battery life. 20 hours of battery life on media consumption, on a laptop that thin?
Just impossible with current, and almost certainly future, x86-64 CPUs from Intel or AMD.
Microsoft, for instance, has already partnered with Qualcomm to produce ARM CPUs for their own ARM based laptop/2-in-1 (https://www.microsoft.com/en-us/surface/business/surface-pro-x/processor). Now, Qualcomm is so far behind Apple it's not even funny. But, unlike x86-64, ARM actually is a very open ISC. Practically everyone has a license. They're cheap and easy to obtain.
Once Apple proves this is a successful possibility, ARM CPUs by other companies (Samsung, for instance, which already makes the Exynos line) are going to start popping up for laptops.
We'll see what that 20 hours really amounts to, but it seems like far more than you'd ever need? You have to sleep sometime. For most people I think this rounds to "more than enough to last all day."
But there are plenty of other ARM chips out there
The problem right now is that Apple is the only company making ARM chips fast enough to compete with Intel in performance laptops. Qualcomm's flagship ARM SoC gets a little more than half the single-core performance of the A14. The M1 is likely significantly faster, needing active cooling in the new MacBook Pro.
Is this a good or bad thing for people who like Macs? I always feel super confused myself as that brand of computers never felt like it was for me and the cost of them is kind of high (plus from what I understand its messy to reinstall them now with the new chip) - AND I've always wanted someone who likes Apple PC's to go "Yup this is a good thing" or "This is generally bad" and just interpret it in to shorthand
Cheers! Its news I want to have a loose grasp of, but really don't understand you know so helping clearing it out is good.
I would say... good thing. As the average mac user, your new computer is going to be faster, it's going to have waay longer battery life (17-20 hours), and it's evidently going to be the same price.
There's really not much downside for the vast majority of mac users.
This turns Macs into glorified iPads, which is probably exactly how Apple thinks of them. I remember how happy I was when Apple introduced Intel Macs. Suddenly, Macs were just PCs with a better operating system. Now they’re back on track to lock you into Apple world completely, even desktop/laptop models. I’m sure their endgame plans for all software to be locked into the AppStore, too. That Chip sure has a lot of “security” features.
Some else make a good desktop OS, please? Anyone who isn’t google?
I don't see how this is true, and it seems like kind of an overreaction. If anything, I was underwhelmed by the amount of changes made here. I also don't see how an internal architecture change that really isn't accessible to end users "locks you into the Apple world completely"? You couldn't service the MacBook RAM or replace the Intel chips in previous models, what's changed here?
MacOS BS still ships with all of the same security features as Intel-derived macOS releases of years past. You can still disable and enable SIP. You can still boot into various recovery modes. Is there something fundamental you're expecting not to work in this new system that did work in previous models?
There's lots to lament about this transition but honestly this seems like the wrong area to be pinging Apple on.
Well, ACKCHYUALLY, you could replace the RAM in Intel MacBooks up until some point (I remember doing that, you had to screw off the back plate and literally slide it in, it was plug and play). So yes, this started before their move to ARM but it's kinda the trend line I'm talking about. It seems like they're closing the Mac ecosystem down, gradually and have been doing so for years.
Also my iPad comment – while definitely referring to the trend more so than the current state – is based on the recent macOS changes which, for example, allow iOS apps to run on Mac and changed the UI slightly to fit the iOS design better (which makes little sense for a mouse-based UI, btw).
People don't like being pessimistic about things that have no alternative and I get that. But I certainly don't feel like I'm being delusional or overreacting when pointing out this trend. Apple clearly hated how Macs weren't integrated enough with their iOS cash cows and plan to close that gap. I don't really see a scenario in which this doesn't end up with a merger of macOS and iPadOS over the next 10 years or so. And the question then is: Among the incompatible parts, which side do they choose?
do you mean MacOS and iOS? because iOS and iPadOS literally split last year. And apple emphasised that they're not planning to merge both OS. I mean I also don't see them ever locking MacOS down like iOS, this would just move developers away from the platform for sure...
This line of argument has been around for at least as long as iOS. Apple has had ~10 years now to “lock down” macOS like iOS. If they were going to do it, why wait a decade? Federighi already mentioned at WWDC that Big Sur on ASi will not be any more locked down than on Intel. The only real big lock down is that non Apple OSes will have to be virtualized. If you are running macOS, it’s still macOS. Install your software from the App Store, from .pkgs, from binaries off homebrew, or compile from source. Same old same old.
This is only really true for non-power users, and is a great security feature for grandma. Every barrier can be disabled, with the intricacy and steps required proportional to the security risk. Having the barriers in place make it harder for grandma to install malware.
Running an unsigned app is as simple as holding the option key the first time you open it. I’m really having trouble seeing the problem here. That trade off seems more than worth it.
This is a slippery slope argument that is a bit far for me. The fact is that unsigned software is, as a rule, going to be less vetted and less reliable than something that has a dedicated developer maintaining it and taking its use in the Mac environment seriously. People using such software should be wary and it should be gated behind some kind of impediment to make sure that only people who know what they’re getting into do it.
I don’t know what’s “predatory” about pointing out “this app may be insecure” when you’re opening an app that hasn’t gone through the hoops to verify that it’s not insecure.
It’s a statement of fact, not sensationalism. And if you’re someone who would be scared off by that anodyne message, you’re not someone who should be installing random apps off the internet in the first place. Seems like it’s working as intended to me. And, again, holding down the option key while opening the app one time ever is hardly a “pain in the ass.”
Can you name me a software that is not “Apple-approved” that was developed for macOS that you had trouble running? I have been using macOS as a daily driver since 10.4. I’ve never had an issue. If you want to go muck about in the system with SIP disabled, that’s one thing. But, just running stuff as a normal user or a developer is really painless in my experience. Hell, if anything, the maturity and stability of the macOS platform has only seen increased software support over time in my experience. Tons of software that might have been Linux only 15 years ago is now easily accessible on macOS.
How is the ASi transition going to change anything on this front, though? If you are already having issues with legacy apps/kexts that are not signed, IMO that is on the developer of that software. I don’t see how it’s Apple’s fault you are reliant on abandonware. And I don’t see how the M1 changes this situation (for better or worse).
Sorry, I really didn’t mean to be malicious! You seem heavily emotionally invested in this so I apologize if what I said came off as mean. What I’m trying to say is that, if developers need to abandon macOS because of Apple’s developer program costs, then the users of that software should move away from Apple’s platform. This is true of any software—if the developer stops supporting it on your platform, then you need to either 1) find an alternative software or 2) get off that platform.
But it’s not locked down for users. It’s locked down for developers unwilling to sign their apps for whatever reason. If you want a truly open system, you never would consider Apple platforms to begin with. I’m trying to understand where you’re coming from, but I just am having trouble because you seem to have strange expectations about Apple and how they operate.
Tell me they haven't spent the last 10 years moving exactly in that direction.
I’m telling you that they haven’t. If you bought a Mac in the past 10 years and it was within Apple’s support window for software and hardware and you kept up with OS updates and the software you used, I still contend that there was no appreciable “locking down”. If you were a developer for macOS and did not keep up with Apple, that’s not true. But that has never been true of developers for Apple platforms. So, I don’t see this as a real sea change. Just an inevitable reinforcement of the walls of their walled garden.
macOS is easy to program for and easy to install programs on. What is not easy (and is getting harder to do on all proprietary platforms, industry wide, not just macOS) is installing arbitrary programs from sources that the platform nor the user trust.
If the user does trust the source (but Apple doesn’t) then the user can still get through Gatekeeper, they just have to do a couple extra keystrokes/clicks. I don’t see how protecting those users who don’t know how to protect themselves by default while maintaining the flexibility for power users to run untrusted software is so draconian that we have to resort to hyperbole about macOS being “locked down like iOS”. On iOS, I’d have to do a lot more to side load apps to get around Apple’s approval process on the App Store. It’s not even close.
The original claim was:
This isn’t true. If you want to believe that Apple adding security features to macOS somehow turns it into iOS, or that Apple was somehow waiting 10 years to pull some malicious bait-and-switch, I can’t stop you. All I can do is try to refute these claims with reasonable evidence.
Fair enough. I may have assumed you agreed with the original position due to the context when it wasn’t appropriate.
I have... I don't feel like there's a lot of respect for UX design in free software and the task of getting that extra level of polish and compatibility seems too big for non-commercial software. I want there to be a free and open alternative so bad, but I currently don't see any.
In what way? I'm very confused. What about personal computers is married to x86?
To be honest I really don't see the difference. It's not like ARM is Apple proprietary. It's less proprietary, even, than x86-64 considering all of 2 companies have a license for that. And it's not like you can run other x86 software without recompilation on MacOS, since the system calls, kernels, and so forth are already different. Software that does run without recompilation (i.e Java, good ol' Electron apps, other things that have their own runtime, etc) will do the same on ARM.
What changes? Nothing stopped Apple from disallowing third party apps, that's a software thing. And they didn't for intel, maybe they will in the future, but they could've done that anyway.
While the loss of bootcamp is sad for developers, with Windows also pushing for ARM I wouldn't be that surprised if we saw bootcamp for windows back in the future.
But to be frank, this is extremely niche. If you're buying Apple hardware to run another OS... It's probably not the most efficient way to do things.
Neither of these make the Mac "just a glorified iPad". Boot camp and hackintoshes don't define the Mac "experience"... Like at all considering the latter is explicitly not on a Mac to begin with
Well, bootcamp kinda did. That’s what made the mac so sexy in the time of the rise of the geeks. Even if you were never using it, you had a sense of freedom knowing it was there.
From a marketing perspective, causing nerdy influencers to get vocal about how awesome it was really raised the mac’s profile among the general public.
For the majority of the market, I fail to see how any of this is relevant. Only an extraordinarily tiny fraction of people—I cannot emphasise how tiny—care about any of this. Most people just want to get on with their lives and create, work, and do their job.
Most people who care about "open" and "free" aren't in the Apple ecosystem anyway. So why is there always hand-wringing/concern trolling about this sort of stuff when Apple makes a change that—let's be honest—probably doesn't have any meaningful end-user effects?
This is a strange point to make. Of course we're responding to how it fits our own needs, not making a judgment for if it impacts enough users for it to be a bad business decision or whatever.
Why is it strange? It's not bizarre at all to discuss whether there's merits to concerns being raised about a product especially if you're not part of the target market. Of course you're allowed to make judgements on them—although usually such judgements wear thin on Apple enthusiasts, there's nothing quite like an Android phone owner telling us why iPhones are bad for the trillionth time—just as I'm allowed to rebuke or question the validity of a judgement because my perception is that such judgement isn't really valid considering the intended purpose of the product.
technically 3. VIA has one too.
Itt-even tildes, and myself included, are not immune to the apple pc wars...
Good clean fun compared to the recent political actions.