Teardown of the 14″ MacBook Pro M2 with Apple’s Help
ifixit.com
ifixit.com
From time immemorial, Apple's new releases have only ever had incremental performance gains from the previous release of whatever model. It has always been this way, and it is true of the M1 and of the M2. What is mind-blowing about M1 is its efficiency, and that Apple switched platforms seamlessly while still incrementally increasing CPU performance. I don't understand why anyone expects things to change and to see a massive increase in performance at every model refresh when that has never ever happened. Yet at every new Apple release, there's always this disappointment from pundits and people who should know better that the new machine isn't exponentially more performant than the last model. What is causing this?!
* It uses a fabrication process that noone has gotten their hands on yet. These have always been massive jumps, and giving access to that node to AMD and Intel gets them to similar performance (actually, they already are at similar performance with the same gen).
* You're buying an un-upgradeable SoC. Soldered on everything means fast interconnect, while others have to play ball with standards that allow me to change components whenever I want.
* It's a pretty damn good CPU.
So, out of these three, Apple is responsible for 1/3. Dump an i9 on a SoC with a 3nm process and it'll eat the M1 alive. There's no "fundamental" increase.
un-upgradable SoC is a strategic design choice so which is basically part of the third reason which is their responsibility as you pointed out.
So basically their success with m1,m2, etc is a well done implementation of a very ambitious strategy to disrupt the market. I don't see why it should be disputed.
I don’t care. I’m impressed.
I don’t really care about benchmarks, these things have allowed me to do twice the work with half as much pain. That’s not incremental.
Maybe they’re not great processors and it’s just Apple cheating in software/process-node/whatever. Great! Let me know when other manufacturers figure out how to cheat in software/process-node/whatever and I’ll consider them.
However, they did just force you to buy a new $2000 machine next time you want to upgrade in 3 years because it's a single, monolithic block.
>Let me know when other manufacturers figure out how to cheat in software/process-node/whatever and I’ll consider them.
Unfortunately, they all accept to be part of a greater ecosystem that doesn't attempt to fuck you over by being un-upgradable and incompatible with each other, so cheating is out of the way.
They should do that!
> they did just force you to buy a new $2000 machine next time you want to upgrade in 3 years because it's a single, monolithic block.
You’re absolutely right. I love my really fast, cold-running, forever-battery monolithic block. I’m very happy to pay ($2000/365/3 = $1.82) per day for it, minus the resale value it’ll still have.
I understand the value of the ideals regarding end-user upgrades, but at the end of the day the tradeoffs just don’t make sense for me.
No it didn't. Look at the Mac mini benchmarks.[0] It is a smooth curve of increasing performance from the oldest models up to the M1. Performance is increasing faster, as should be expected, but the increases in performance from model to model are incremental, always resulting in "experts" being surprised and disappointed. M1 Mac mini is not ten or five times faster than the 2018 Mac mini, and not even twice as fast. It is 26% faster in multicore performance and 36% faster in single core performance. These are increments, not massive leaps in performance.
I do agree that Apple's CPU releases are incremental updates (and mostly always have been), but when I used an M-series Mac for the first time it was a step grade upgrade from the previous gen even if it's somehow not accounted for in CPU performance metrics (though I think it will be if you account for energy as well).
Since when? Seriously. Since the first Apple Silicon release, undoubtedly.
My laptop doesn’t heat up anymore. It doesn’t die. If it has fans, they’re absolutely silent. All this, and I haven’t had to change my workload at all.
I don't think that's really true.
The major cause for the failure of the macbook was a lack of power efficiency. Intel just didn't make a performant enough low power chip to make the concept work outside of a small group of people who value portability over everything else.
But even beyond that, there have been frequent complaints of Apple's laptops running hot. Those complaints don't necessarily show up in benchmark numbers, but I've run across them many, many times.
In what world or category did the Macbook fail? It's consistently top selling and top rated.
Yeah - sorry. I should have said "12 inch Macbook". It's unfortunate that Apple's naming in that instance is so confusing.
Prior steps were about 10% or less.
Example: On the last Intel MBP I could barely run Teams with video, the Intel Macbooks (I tried many) got immediately super hot and started throttling to a point that made the machines unusable. The M1 Max doesn't even turn on the fan.
What you have done is leap frog a couple generations there. Compare the benchmarks[0] of one model of the last Intel MBP, say the 2019 13" MBP (MacBookPro15,4) and the very next generation of that model, the 13" 2020 M1 MBP (MacBookPro17,1) and you will see it is an incremental increase in performance, a 36% performance increase in single core and a 40% increase in multicore performance. Impressive, but these are not exponential gains in performance nor even a doubling of performance, like what everyone seems to expect between the benchmarks of M1 and M2, and in fact this foolishness is not new, and has been going on since 68k models were new, all through the PPC era, into the Intel era up to today.
[0] https://browser.geekbench.com/macs/macbook-pro-13-inch-retin...
https://browser.geekbench.com/macs/macbook-pro-16-inch-late-...
https://browser.geekbench.com/macs/macbook-pro-16-inch-2021-...
And the real world usage implication feel even more crazy than the numbers look like.
Also, even the one you linked looks crazy. Almost 40% increase - when did you ever get something like this in a year to year upgrade?
Additionally these benchmarks don't really tell about the thermal throttling problems the Intel machines had, which are completely gone with Apple Silicon. So all in all I'd definitely not call the Apple Silicon jump incremental - for the Apple hardware it was revolutionary.
And let's be honest — synthetic benchmarks are bullshit. In this thread, you have a number of different people describing their experience of how much the perceived performance gains were, and how they didn't feel incremental. You are bending over backwards to try and dismiss those, and I don't get why. My perception, between having a maxed out 16" Intel MBP and getting an M1 mini at the time, was nothing short of "holy shit, this thing smokes my $4000 machine". Call it incremental; call it whatever feels right to you, but I know what it felt like. My sample size of 1 analysis is: it was not incremental at all.
(An aside: M1 to M2? definitely incremental.)
Coming from the lower end of the product line, the M1 was an incredible upgrade, and the price point really didn't change (I paid 997 for the Intel, $949 for the M1). I swapped an Intel Air for a M1 Air, and could build Go and Node apps in 25% of the time it took the Intel Air... and could do so all day on battery.
iPad 2 a year after iPad 1 had 2x faster CPU, 9x faster GPU, 2x RAM, weighed 15% less, was up to 33% thinner in every dimension, and had longer battery life.
[1] https://www.notebookcheck.net/SGX535-vs-SGX543MP2_2376_2572....
Also, how does the stuff about DDR vs DDR2 RAM fit into this conversation at all? What you're responding to seems to be a claim about RAM capacity rather than speed.
The change from in order to out of order was huge. L2 cache doubled to 1MB. Memory bandwidth doubled. The FPU got properly pipelined too. RAM also doubled to 512MB.
In combination, this translates into 20% better integer performance and double the FPU performance. This is pretty close to what other chips saw moving to A9.
For stuff like web browsing, that FPU update basically doubled performance and made the iPad 2 much more useful to consumers.
On the GPU front, Apple themselves made the 9x claim, but Anandtech showed closer to 6x. That’s still a massive improvement. If something was limited by the previous 256MB RAM, that 9x number becomes very believable.
You are comparing apples and oranges. Within specific model releases, MacBook Pro vs next MacBook Pro, Mac Pro vs. next Mac Pro, and every single other Apple model release compared to it's previous and subsequent releases, the performance gains have always been incremental. I don't understand how anyone could expect it to be otherwise.
The parent is arguing that really, these are the same product category. If you bought, say, a middle-of-the-road MacBook Pro every time a new one was announced, you’d see a huge jump in performance (more than just incremental) when moving from Intel to M1, then a return to incremental improvement from M1 to M2.
I’m pretty convinced by them. I think that Apple has done such a good job of compatibility that all MacBook Pros are essentially the same product category. And the M1 performance improvement really was a step change over Intel.
> gains over intel 9 Macbook Pro on the higher end Mac Pro's
I'm saying you can't compare a MBP to a Mac Pro, like you tried to do, because that wasn't my claim. My claim is that the first Apple Silicon MBP was only an incremental increase in performance from the last Intel MBP. And I am saying this is true, within specific model lines, for everything Apple ever released. And I don’t have specific knowledge of this, but I would assume it is true for every other computer and platform, that each subsequent generation of model is a little faster than the last, and not twice as fast or even 50% faster.
Do you have anything to support this claim because all the benchmarks I’ve seen say something quite different.
The first Arm versus the last Intel? Base model versus base model. It’s an awful lot more than 35%, particularly in multi core. And it’s quiet and cool. It’s a massive generational leap.
They're fundamentally different architectures, so when the Air experiences the jump in performance that it did, you can't just hand wave that away because "well if Intel had smaller and cooler cores, it wouldn't be as big of a difference".
On the contrary - this is the entire point.
Anecdotally, my 1st gen M1 air is significantly faster at almost every task when compared to my last-gen i9 16" MacBook Pro.
It seems your argument is just that some increments are smaller than other increments? It seems the primary point is that the move to M1 was a much larger jump than most jumps.
Anyone who has been purchasing Apple hardware over the years experienced this in a very practical sense.
I’m looking at a Geekbench of 1707 for the 2020 M1 and 1113 for the 2020 i7, 1048 for the 2020 i5 and 972 for the 2020 i3 (as shown in the link I provided). That’s an increase of over 50% in single core performance versus the i7, more than 60% versus the i5, and 75% if you compare to the i3.
Am I looking at the wrong machines? How are you comparing?
M1 MacBook Air was 2-3x as fast as previous model. MacBook Pro was closer to 2x.
Both with dramatically improved battery life.
https://www.tomsguide.com/news/macbook-pro-m1-benchmarks-are...
Incorrect. The Late 2020 M1 MacBook Air has just under a 35% performance increase in single core performance over the Early 2020 i7 MacBook Air, though nearly a 2.5x increase in multicore performance[1] (because it has more cores). That doesn't make it 2-3 times faster as the single core performance is a bottleneck, but its multicore performance increase is certainly impressive, it just means the Air's small formfactor was hamstrung by Intel multicore and got to breath a little in the platform switch with more cores, not that the last Intel models are 2-3 times slower than the first Apple Silicon models. That's absurd. The MBP has much less impressive performance increases in the platform switch compared to Air. Looking at the last and fastest intel MBP, the 13-inch mid-2020, and comparing to the first M1, the 13-inch Late-2020, the performance increased[2] about 30% in single core performance and increased 40% in multicore performance, which is typical of the performance increases between the last Intel and first Apple Silicon selections of most other Mac models.
[1] https://browser.geekbench.com/macs/17
[2] https://browser.geekbench.com/macs/macbook-pro-retina-intel-...
M1 vs intel looks to be a 1.5x to 2x improvement on the air in synthetic benchmarks.
MBA 2020 (early) i3, i5, i7 are scoring 972, 1048 and 1113 respectively.
MBA 2020 (late) M1 scores 1707.
thats 175% of 972 and 150% of 1113.
Am I wrong?
Thats a huge leap in performance for the high end, but even more so for the entry level.
Yes. And you're making it overly complicated. Look at the benchmarks[1] for one single model, the 13" mid-2020 MBP, single core performance is 1213. Now look at the very next generation of that machine, the 13" late-2020 MBP M1, single core performance is 1708. What percentage of 1213 would you need to add to 1213 to make it equal 1708? About 40% of 1213 then added to 1213 would give you 1708, so the increase in performance in single core for this model jumping from Intel to Apple Silicon M1 is a 40% increase, not 150% or whatever.
[1] https://browser.geekbench.com/macs/macbook-pro-retina-intel-...
This benchmark shows the difference in performance of the M2 devices when you run them for a long period:
https://wccftech.com/m2-macbook-air-throttling-problem-under...
The success of the M1 is not just raw performance numbers but its efficiency.
This is where we disagree. I see a smooth slope of performance increases generationally. You see some kind of chasm opening up.
And you still haven't explained where you get 'less than 35%' increase from.
In other words, drumming up excitement/disappointment is a feature, not a bug.
(Otherwise factually I agree with everything you said)
Pundits criticize, that's what they do. It's impossible for anything to ever be "perfect". If a product is simple then it can be criticized for not having enough features, if it's full of features it can be criticized for not being simple enough. And when everything's pretty much fine, the standard criticism is "the new version hasn't changed/improved enough". (While when lots of things change/improve, it becomes "but is it too different/incompatible for people used to the old version?!")
The job of a pundit isn't to try to provide any kind of objective or even consistent truth. It's just to provide talking points than turn into headlines that drive clicks.
The pundits, as people, absolutely know better. But that's not their job as a paid pundit. Their job is to drive clicks and comments and engagement.
We all have different financial situations. Tell me what’s good and bad about a product and let me decide if it’s “worth it” for the price point.
It's not disappointing, but it's hardly impressive. Once again; this is the world's largest company, and we're freaking out over TDP trades for performance. It reminds me of the people in the 90s who would shill Disney's theme park tech; maybe it's cool on a technical level, but these people have billions of dollars. What's impressive is running Linux on unsupported hardware or single-handedly getting a copy of the no-fly list. I can't be bothered to give a fuck about Apple spending 18 months to make a 40% faster GPU that consumes 20% more power.
That's not irony. Irony is that, right now, no other hw manufacturer can compete with Apple. I don't think Intel or AMD will ever be able to compete again until they abandon x86/amd64 and release their own ARM chip. Apple will not sweat before that happens.
Apple made the right choice abandoning Intel on the 10nm node, but they don't have a roadmap from here besides "get better silicon". The competition is hot, and I don't think either my next laptop or desktop will be ARM based (unless someone out-performs Apple).
Right now, AMD is showcasing a processor, the Ryzen 7040, and proudly comparing it to last year's M1 Pro using cherry picked benchmarks. That's how its going to go from now on. They have nothing to compete with M1 Max or Ultra, or Apple's current flagship processor, M2 Max, and by the time they do, Apple will have left the M2 Ultra behind and will have the M3 Max as their flagship processor. That isn't competing, that is chasing.
If we're only getting more 20-30% spec bumps every 18 months from Apple, it sounds like the race is pretty close.
So there is no point releasing a major update and making the M1 look obsolete. They just need a solid update.
In time we'll see if there are greater improvements coming. Something tells me that we're never going to see a performance leap comparable to the 10nm -> 5nm one, though.
The whole approach centres around consumers feeling like if they do purchase a cheaper M1 model that it is a great, future-proof device and not an obsolete one.
To Apple, marketing and product strategy are deeply synergistic. It is for every company at their scale, and especially when you have to sell something direct-to-consumer.
And the only people buying them were those that needed Intel apps that weren't yet optimised for M1.
Not that those comments can't be on point, but it feels like everything now has something like that.
And yeah efficiency is the impressive part of those chips.
Reviewers online aren’t particularly well informed. They think they are because many of them sort-of-kinda understand the devices they’re talking about at a high level, but they’re not really familiar with how it works and what the constraints are.
M2 has very predictable performance, on a line with all other CPU gains in recent years.
M1 had a lot of low hanging fruit to reach for that standardised architectures aren’t able to mimic for fear of breaking things (co-located memory, purpose-built buses, function-specific co-processing). And then they got on a node no-one else could get on due to their backlog.
I watched that kinda cringe LTT mea culpa video over the PS5 storage architecture a while back and I was a little taken aback at how much of the basics he didn’t really seem to understand, even in his apology video.
These very bespoke systems can take some really innovative steps with custom compression hardware and expanded bus space.
It’s not Intel or AMD that can’t keep up with Apple, it’s the standards they build their hardware on.
Now here is your contradiction:
> Each of the M1 products has been a big step up from the preceding versions in almost every respect and especially performance.
> ARM architecture by itself is not significant performance advantage or competitive advantage at all, and there's no real secret sauce in Apple's chip program other than tight focus, a long time horizon, and massive amounts of cash. And CPU performance growth is slower than it once was as anyone who remembers the road from the 286 to the Pentium Pro can attest.
These statements conflict. I agree with your second statement. Regarding Apple Silicon, the leap forward was in power efficiency, in the platform switch, which by itself was pretty amazing, the performance increases are NOT 200% increases across the board on all models. Apple Silicon is roughly a third faster than the previous Apple Intel model generation, up a small amount of increase from the previous Intel generation to the last, and if you go back through the models you'll see these performance gains have been increasing, but the gains are incrementally faster on an incremental performance gain at each generation. This shouldn't be surprising. What is surprising is you and everyone raving about the massive performance gains that aren't there. There is a performance gain, but it is not earth-shattering, like if each gen was performing twice as fast as the previous. They're not, and they won't, if they ever do, for years if not decades. The refreshes are always incremental, Apple Silicon is no exception.
As for the second, that's not a contradiction. ARM ISA has very little to do with Apple's in-house processor performance. If they had stuck with PowerPC they could have achieved roughly the same thing (on a technical basis, licensing is a whole separate issue and likely the biggest driver of the switch to ARM).
Are you expecting AMD (or anyone else) to compare their new processors to things the competitors have not yet announced?
What would make Apple sweat at this point? Google must have when it seemed like they were going to produce high end consumer versions of ChromeOS laptops and Android phones. That's the kind of integrated experience that would compete with enough money behind it to get somewhere. Google blew that one though.
I'm not like pro Apple, I just don't think they care that much what Intel does.
Backwards compatibility can be nice, but if implemented backwards it starts hurting your hardware development (and software adoption) pretty bad.
As for reviewers/writers having bad takes, this keeps happening as long as they don't understand the goal that things were designed for. It's easy to always assume that big bad hardware corp made it so that you get the worst possible product for the highest possible price, but if we just take hardware pairing for a moment, this gives you two things: zero-trust guarantees ("just because the hardware is plugged in doesn't mean we trust it") and actual production (you have to actually have the keys to decrypt anything or prove ownership to the system, it's not just some pinky swear). As a downside, as long as PKI is the least-worst option we have for doing this, it means that Apple fully controls this (same as Intel with ME/BROM and AMD with BROM and AGESA), for which nobody really seems to have an alternate solution yet. Luckily for the big hardware companies, most users don't care and aren't really affected by it either. Unlike tractor repairs (those things tend to break frequently enough to warrant end-user expertise), the overlap of people that are in need of repairs and the people that have the expertise to do it is extremely thin.
The performance has never increased by the same amount year over year. It improves drastically in some years, and modestly in others. People get disappointed when Apple fails to deliver the performance gains they have in years past.
Edit: Also the [sic] is unnecessary. Just say you don't like the publication if you disagree with their writing that much.
Year over year is never a huge bump, not in features or power.
But if you're upgrading a 5 year old device, the bump is unbelievable.
I don't know how much of this is due to the new regulations, but it has been a very smooth process.
https://www.bar.ca.gov/arsc/newsletters/newsletter/fall-2021...
I’m not convinced Apple came to the conclusion that it ought to do this for business reasons.
Just last week I upgraded the memory in mine, no repair kit required, just a simple screwdriver.
I've used a (presumably leaked) 2008 MacBook Pro 15" service guide to disassemble, mend the bent case (due to a bad fall), change the hard drive with a SSD, and put my machine together.
I think this is unfair, because precisely what the OP is saying is that in e.g. Lenovo products these instructions have been readily available for customers, sometimes even in the very product manual itself. I have an HP tablet from around 5 years ago where the disassembly instructions came with the printed manual. Also, it requires about 3 tools only, all of which are so standard that except for a suction cap I think they are almost in every household. Compare to this teardown of a much thicker laptop... even for replacing just the battery, I wouldn't have the tools, much less the balls to do it. It's not a surprise that iFixit still gives them a rather low "5" score.
I just say that these guides are always produced, but not generally made available to consumers, and that's bad.
I wish every electronic equipment is shipped with a User's Guide and Service Manual out of the factory.
However, both to protect service businesses and due to more integrated nature of newer devices, self-service is becoming more impractical every day. Hope right to repair changes that for the better.
Also, there was an Ask HN about "Declination of Everything", and people objected vehemently. Why we are having this debate now, if nothing is declining?
> There is of course the elephant in the room: parts pairing. As it stands, despite the increasingly repairable designs, the software locks that Apple maintains will result in waste as otherwise-useful components end up in landfills instead of being repurposed. The useful life of our devices will also be limited to Apple’s hardware support—whatever they decide that may be. Once support is dropped for a device, those software locks will remain in place which means even if a third-party manufacturer is willing to step in with replacement parts, those parts may be restricted in functionality.
> It took us 20 years to get these manuals ...
An 16 GB MacBook Pro can live as a development machine (at least for what I develop) for almost a decade, so getting a 64G machine today will not make it obsolete for me till the machine itself dies.
Also, an 16GB MacBook is pretty enough for running a lot of applications, incl. Electron based hogs. MacOS manages its memory pretty well.
Neither RAM nor SSD is embedded in the SoC. The RAM is just soldered on the package. While this makes for slightly shorter path lengths, the only real gain is motherboard footprint.
As an old overclocker, I don't believe that this compact soldering is only due to motherboard footprint. It would allow them greatly push the RAM chips to their limits.
Well, that and the ability to use LPDDR5 in the first place. There’s no such thing as a pluggable LPDDR5 stick; the design decisions around low voltages & high signaling rates make it physically impractical to survive the longer traces and discontinuities involved in the signal traversing a connector.
You can solder it on the motherboard, you don't have to solder it on the package.
And because this is soldered down, you can’t fix this like how it would be doable on other laptops.
On the other hand, I still think that having slots adds a lot of bulk to system design. Considering SSDs have limited life spans, being able to change them makes great sense, but when you dedicate 90% of your footprint to battery, it becomes hard to make them replaceable.
One might argue that HP, Dell and Lenovo offers this, and I suggest buying these machines if that's importance to you. I'm fond of all three companies' higher end designs, but Apple provides more at the same price point, at least for me.
This is really nitpicking.
Random I/O is the number most people are affected by and there’s a reason the clickbait commenters never mention that: you don’t see this large an effect and these machines are all more than fast enough for the vast majority of buyers. That will not get you a ton of YouTube impressions so it’s all about a use-case these machines don’t have.
While their absolutely is benefits to how they handle their RAM the upgrade prices are nuts. $200 extra for 8GB more, 64GB requires a CPU upgrade and costs $400 extra over 32GB. I've heard the CAMM modules Dell developed and seem poised to replace SO-DIMMs in laptops will allow for faster speeds and replaceable LPDDR too with less Z height and eaiser board design so hopefully we will see an uptick in PC laptops with replaceable memory again.
I don’t know why manufacturers charge so much for spec bumps (this isn’t limited to Apple). Is it really only because they can?
This also played into sale or discount strategies: if the same model becomes cheaper over time but list pricing doesn’t adjust, you’re making a good bit more from everyone who isn’t price sensitive even if you allow resellers to discount it periodically or offer better pricing to business orders (“free RAM upgrade”).
This is addressed in TFA, which suggests it's a side effect of smaller 128GB modules becoming scarce — or maybe more accurately, "not reliably available at Apple volumes".
The article asks "why is Apple even bothering with a 512GB version?", and the two answers are: (1) To create a cheap-ish "price anchor" config, and (2) as a reasonably-priced choice when price is paramount.
Also, for those who want to do 4K video editing or whatever, it's easy to throw a couple M.2s in an external dual Thunderbolt enclosure¹ to create very large, very fast local storage to complement the 512GB boot/apps/documents drive.
¹ https://www.bhphotovideo.com/c/product/1635760-REG/sabrent_e...
Have you ever done supply chain and/or pricing work? Nothing like this is ever “just”. Especially at Apple’s volumes.
I can’t off the top of my head think of any CPU that’s ever supported running different DDR standards side-by-side simultaneously (there were some dual DDR2/DDR3 boards back in the day, but it was an either/or proposition).
It's cramped inside, so no go.
*I’m no expert just a wild guess
It's pretty clear that Apple's decisions about manufacturing approaches are not to make iFixit's job harder or easier, but to avoid needing iFixit at all for most users.
The connector exceeds what even the fastest NVME SSD can deliver and the form factor of enclosures are small enough that you can simply carry it everywhere.
(Actually they spun off their canned food division a few years back under the name Ace of Diamonds, which was then snapped up by some Thai company for a song. Nevertheless...)
Japanese keiretsus are weird, man. In Japan, 7-Eleven does banking.
We get it, you don't like companies that don't make the most repairable products. It gets a little tiring
Whoever was on this team at inception and whoever is still running it today - you are an ember of what made Apple, Apple. Thank you sincerely. Truly, you all should be expert witnesses to the antitrust cases to show what exactly it takes to do what everyone is expecting to ‘happen overnight’ done excellently.