I don't understand the comparison to the iphone and android situation. As far as I can see, android phones are highly competitive with iphone on everything. The longest battery isn't on an iphone, the best camera isn't on an iphone, the latest screen tech isn't on an iphone, etc. etc.
Coming to the M1 laptops, they beat comparable Windows laptops in every single metric. It's literally just a matter of looking at the numbers and realizing how M1 has every Intel chip beat (in this category).
In fact this makes me wonder: if Apple chips are so far ahead, why are iphones not the fastest, best smartphones by a wide margin?
I don't know about the other metrics you talk about, but iPhones _are_ the fastest. In general, I find iPhones better than Android for general use - but that's a subjective opinion. Apple provides the best overall package, even as the only thing they beat everyone else on is speed (to the best of my knowledge).
I think this is often about the experience of using it as opposed to raw compute performance benchmarks (even if M1 does excel in those). In many cases, be it because of integration, more care about design, or whatever, the apple experience feels more... snappy? Smooth? It's hard to put a finger on what exactly leads to this, and it may also be about the combination of many small details.
Snapdragon 8x5 CPUs have been fast enough for a while now, and you wouldn't notice a difference as long as you run a decent flavor of Android. However you can notice a faster screen refresh rate, and high end Android wins there. I wouldn't trade 90 (or 120hz) screens for a faster CPU.
I think it's a bit of both... I don't doubt for a second that Apple creating both the CPU and Software together are why the X86 emulation is as good as it is.
I'm actually thinking about getting one of these as I've been without a personal laptop for a while now. That said, I'm waiting for Docker support to get flushed out a bit more as I wouldn't mind getting work done. I am interested on how the really high performance ram effects my workflows, as I've often gone over 16gb as a sticking point, got my work laptop upgraded to 32gb earlier in the year and my personal desktop is at 64gb.
I'm not sure it will replace my r9-3950X desktop for getting things done, but it's absolutely compelling for its' space. I swore off macs for a while now, and this might pull me back in.
I have used the latest iphone and the latest galaxy note and honestly the note feels faster. Maybe because of the screen refresh? But then the same question: what use is a faster chip when the end result feels slower?
Plus they seem to last a lot longer than your average android smartphone in terms of longer term performance/updates. I'm holding out a (faint) hope that they'll make at least one generation with USBC before going port-less as that's the main thing stopping me from switching
Overall experience feels better on android. Few things that make difference for me.
1. Notifications are leagues ahead of anything on iOS.
2. Little accessibility features like sound search, automatic caption, text selection from any screen/pictures, better integration with google assistant adds up.
3. Customizability. Yes, even now iPhone is super limited in the layout you can have on home screen or the new app launcher. You can't group things or put them wherever you want on the screen. This is ridiculous.
4. Android phones have higher refresh rates. Almost any flagship in 2020. The punch hole camera feels better than the notch. Face ID = problematic in the pandemic with masks. Though, I like it generally but apple could have given touch ID on the same device as well.
5. For tinkerers, it has better support again. Youtube vanced, tachiyomi, advanced adblockers, "real" firefox with extensions, etc are only available on android.
The best part is you can get the pixel 4a for $300 with some carriers. 3 years of updates which is what typical upgrade cycle looks like for iPhone users even if they get "updates" for longer.
You're not wrong, but things like a unified clipboard, continuity, the peace of mind which comes from knowing that there's a very small chance of malware coming in from the App Store, and (again my opinion) more polished indie apps like Ulysses, Things, Fantastical etc. keep the iPhone ahead in my book.
I think notifications are perfectly fine on iPhones. I've not used android for 4 years now, but if their notifications are the same as they were in 2016 (or just marginally different) then I don't miss much. I confess point 2 sounds very appealing. I do use custom DNS for adblocking, so I don't think I miss it on the iPhone, and tbh I've fallen in the rabbit hole of tinkering with my phone enough that I feel more productive without a phone I can tinker with.
No, notifications have changed a lot in the recent years. And you can get unified clipboard by installing app for your OS (windows or linux) through playstore. You should try a new pixel device to see the improvement.
Some apps are better on iOS but it's not a big difference because some apps are better on android as well. People care about different apps in the end.
You're 100% right on notifications -- despite having borrowed a bunch of ideas from Android on how to handle notifications, iOS still doesn't get it completely right.
There was a point in time that I was pretty deep into the Android scene, but at this point in my life I just want my phone to work consistently and have great battery life.
After many years of using Android devices (including many Nexus / Pixel devices), I switched to an iPhone 11 Pro Max last year. I've been very happy with that switch, my phone always makes it through a day of use and I haven't ran into any of the quirks that I encountered on my Android devices on a regular basis.
I completely understand why you would prefer Android, but there are definitely people (including myself) that are happier with iPhones.
Yes. I have the same iPhone too. I just don't get why so many people here are fixated on chipset differences when average people care more about having better quality mic, speakers, etc. Even camera is good enough on most devices these days. The photos will not have huge differences after being compressed on online platforms in the end. How many people actually care about having raw photos?
I just don't think any flagship in 2020 is worth it for an average consumer when options like pixel 4a exists. iPhone se is not good due to the screen but otherwise it would have been a great contender as well. People notice 90hz crisp display more than faster opening apps.
There is quite a significant difference between iOS and android devices, there is no android device that competes with the performance of iOS right now. Heck, most laptops don’t even compete with iPads these days.
An Android could have amazing hardware but everyone knows the OS isn't going to get updated and you're going to have to pay to replace it in 2-3 years.
Lol what are you smoking? iPhone as a whole has trounced Android repeatedly over the last few years. There really is no comparison aside from Android having some [random esoteric] hardware that iPhone doesn't for a particular year. None of this is focused on battery life, performance, and efficiency. It is window dressing on top of an inferior architecture.
Also I fundamentally will not buy a piece of hardware that will cease receiving security updates within 36 months of purchase. That's insane, and frankly a major fuck you to anyone considering buying your product.
You have reduced the scope of the argument. It is a question of efficiency, performance, and battery life. iPhone beats Android in all over those categories.
Of course iPhone and Android can do all the same tasks. One just does it a lot better efficiency and power wise.
How does iPhone beat Note20 in battery life if the Note lasts longer?
How does the chip performance matter if the Note feels faster in opening apps and scrolling due to the refresh rate?
How does efficiency matter if the Note holds more apps in memory for faster task switching?
How does a faster chip matter if the image processing is way superior in the Note so your photos almost always come out better?
The difference is subjective. I would have expected it to be objective like in the M1 laptop. The new air and pro absolutely obliterate other comparable systems in every metric.
The whole isn't the fastest, the biggest comes up with big caveats.
For example the iPhones don't have the longest battery, but the phone that has the longest battery has two times the amount of battery, while having a much higher consumption than iPhones. If Apple decides to stick a 5000+ mah battery in an iPhone - Android phones would have no way of catching up to its battery life.
Another big thing being the latest screen tech - Apple can't afford to do an OLED screen on the iPad, since Samsung and LG can't possibly manufacture enough OLED screens to meet demand. Where Samsung is perfectly fine doing it, when they won't sell enough tablets to matter.
In the end it's not that the iPhone is clearly inferior to Androids, it's mostly that Apple has some numbers in mind that they consider optimal and going beyond those won't increase their sales (certain amount of battery, certain screens, etc).
Can't speak for anyone else, but for me, the biggest sticking point is being able to side/load software and other app stores... not that I pull in much, but I don't want my device mfg tell me what I'm allowed to run on what I buy.
Sometimes this means the software is more clunky, less polished and isn't as fast... but I'm still using my now 3yo Pixel 2XL, which is still mostly acceptable (I really need a new battery).
It also basically doesn't matter for 99% of people. Most people just need their phone to be fast enough to run common applications responsively. Benchmarks are irrelevant to the average user, and are mostly a dick measuring contest. People tend to prefer mobile platforms on the basis of features that they offer, practical hardware traits like battery and connectivity, and familiarity.
Completely agree with your sentiment. But a lot of people are used to thinking about performance like it much matters anymore. You’d be hard pressed to find much of a difference between yearly updates of the same model phone. But I think a lot in this community will look at those numbers, I mean that’s how most people grew up when thinking about PCs and seems to be how Android devices have also marketed to people. I think now the majority of people just upgrade to the latest phone of the OS they prefer. I mean personally after deleting social media and a lot of other items off my phone I wonder what I really use my phone for other than texting, taking pictures and checking email and my calendar. I could easily get by with some basic phone (honestly contemplated getting some e ink phone with basic capabilities). But there are plenty of people out there that use their phone constantly and for a host of things. All that to say I feel like phones won’t have another large advance for some time, small iterative changes. That jump from the original iPhone to the iPhone 3G seemed gigantic. Sure there have been form factor changes but even on the android front there have been changes but not things that are revolutionary to the majority of users.
> In fact this makes me wonder: if Apple chips are so far ahead, why are iphones not the fastest, best smartphones by a wide margin?
Apple has pretty consistently had the fastest SoCs for phones and tablets for a few years now, particularly in single core performance. Here are some benchmark results of the iPhone 12 versus some of it’s Android competitors [1]. Notice that last year’s iPhone 11 Pro Max still outperforms Android phones from this year as well.
I think part of the reason this isn’t a major differentiator (for phones at least) is that phones have been “fast enough” for several years now. Given the very aggressive throttling that is utilized on these types of devices, I think the primary difference ends up being battery life / efficiency rather than raw performance.
It’s also worth noting that while the performance / efficiency gap between x86 and ARM seems to be pretty sizable at this point, the gap between various current generation ARM CPUs isn’t nearly as drastic.
Yeah that's what puzzles me, why has chip superiority not translated into an objectively better phone? For laptops the M1 laptops are best in class in everything, battery and performance.
I think for some people (eg my oldest brother), the iPhone is "the best." He's not a tech person - and that might be perspective that you're missing.
I tend to agree that there are great androids that seem competitive with iPhones (disclosure: I haven't really owned a flagship phone for a while; this is one of android's issues! The flagships aren't that much better than the $300-400 phones)
But I know a lot of people who think the iPhone is the premium phone experience, and if they're going to spend over a thousand dollars on a phone, it'll be an iPhone.
My point here is that for many people, the iPhone is objectively better.
He is being downvoted by people who consider the choice of a specific vendor a religious issue and will vote down anyone who in their eyes disparages their vendor of choice. It is a sad fact that the existence of this phenomenon only leads to the reinforcement of the religious behaviour since their insistence on the infallibility of their vendor is questioned by those not part of their cult, whereupon they double down on the magnificence and superiority of their vendor's products over everything else, which leads to more reactions and the cycle continues.
The power/performance ratio of the M1 SoC does not seem to need religious adoration to be seen as a significant step in the establishment of ARM as a real competitor to the AMD64 hegemony. The wait is now for other vendors - Samsung, Qualcomm, etc - to launch similar "desktop-oriented" SoC's which can be used by both traditional "dekstop" vendors (HP, Dell, Lenovo and to a limited extent Samsung etc) as well as traditional mobile-oriented vendors like Xiaomi. Some of those vendors will eventually produce ARM-based platforms which support user-upgradeable memory, GPUs and storage - and possibly also user-upgradeable CPUs - like they do for AMD64-based systems. Once these systems become available ARM has a real chance to take over the market.
Great points. I remember back in (good lord) 2011 when the MRI research was shown [1]
I'm very excited about the new m1. While I'm tempted to buy it, I also can't wait to see what they do with it in their pro models. Or what the second generation looks like.
The competition is great for everyone, and Apple going all in on arm as they run from Intel's manufacturing failures is very well-timed.
I think this was somewhat inevitable, to be sure. Microsoft had Qualcomm and AMD processors in their surface [2], and I think as TSMC continues to dominate in manufacturing, we will see really exciting gains.
Also, I'm thrilled that CPUs are coming with tensor cores now. I think fast matrix operations might be separated from gpus (at the least, you no longer need a >$400 gpu for it), and that's a win for everyone, too.
In two years, we could be able to reimplement early neural network research on a macbook air! (I mean, you already can, but instead of taking minutes/hours/days, even more recent things become accessible)
To a large extent, current Android flagship CPUs are "good enough" for the things people do with phones. Sure, current iPhone chips are much faster, but it doesn't matter that much for the vast majority of consumers (even for games, mobile games tend to target the lowest common denominator to a much greater extent than PC games; no-one's going to launch a mobile game that runs on the latest couple of generations of iOS and nothing else).
iPhones do likely get better longevity out of their faster hardware, mind you.
> It’s also worth noting that while the performance / efficiency gap between x86 and ARM seems to be pretty sizable at this point
It's really not, though. You have to compare performance at a given power target to judge performance/efficiency, which you can't do for most ARM CPUs. But you can get 8 core x86 CPUs that are 15W, or 1.8w per core (example: the 4800U). Similarly 64 core Epycs are around or less than 3W per core. These are all well within similar per-core power numbers as your typical big-core ARM chip.
The perceived gap is larger than it really is, because there just isn't really an x86 tablet-focused market. The M1's jump into laptop-space puts it up and we can really see it shine, but the other ARM jumps into laptop-space were a joke. So comparing equivalent power targets becomes challenging.
But take your articles benchmarks for example. The Snapdragon manages something like 60% of the performance of the A12's. So take the existing M1 mac, cut the performance drastically, and suddenly it'd be not interesting at all. The performance/efficiency wouldn't be impressive at all if it was getting 60% of the performance it is now.
> I don't understand the comparison to the iphone and android situation.
Because articles like that are pure nonsense.
Take the initial paragraphs. We already knew that faster processors don't necessarily run hotter. People are simply conflating higher clock speeds with higher performance, while ignoring all of the ways CPU design optimizes computation. Likewise, we already knew that "emulation" doesn't imply that performance will be unbearably slow. There has been enormous amounts of research and development into the problem over the decades, to the point where much of our software is optimized at runtime.
None of this is meant to diminish what Apple has accomplished. They were clearly paying attention to details that have been largely ignored or at the very least rarely highlighted by their competitors. Now that it has happened, we will likely see more from the PC industry. Apple will likely have a hard time keeping up unless they have more up their sleeve.
Likewise, we already knew that "emulation" doesn't imply that performance will be unbearably slow.
Apple’s Rosetta 2 doesn't do emulation; it translates the x86 code into ARM code on the fly and runs it, which is why most apps run faster on M1 Macs than they do on Intel Macs--and comparably priced PCs.
Now that it has happened, we will likely see more from the PC industry.
Probably not. If Intel and AMD were capable of this kind of performance per watt, they’d already be doing it. Years ago.
Because the latest x86 chip has to be backwards compatible with 40 years of legacy code, there are certain limitations Intel and AMD just have to live with.
Apple will likely have a hard time keeping up unless they have more up their sleeve.
It's ironic that it's Apple that will have a hard time keeping up when they've already disrupted Intel and AMD. Remember, it's not just a processor; it's a system on a chip that contains an 8-core GPU, 16-core Neural Engine for machine learning tasks and other components.
This has obviously been in the works for several years; Apple wouldn't have gone down this path if they didn't believe it would give them a sustainable competitive advantage for years to come.
Even with its ridiculous performance, the M1 is the entry-level consumer version of the M series suitable for (in Apple's lineup) entry-level consumer hardware. The M1-based MacBook Air doesn't have a fan and it's faster and more capable now than many (most?) PCs that cost more and run hotter.
One thing we know: the M2 will be faster and more capable than the M1 and won't be limited to 16 Gb. We'll probably see M2-based Macs by the summer of 2021.
M1 is still just a tablet SoC in a laptop form factor.
We don't know what the efficiency will look like after you add all the I/O you need for a real desktop, we don't know how the memory subsystem will deal with 4 more cores that are just as fast. What's the point of having fast cores if you cannot feed them?
We don't know how lower yields for bigger chips on the newest nodes will affect their costs.
People act like Apple changed the game or is light years ahead of everyone but they literally haven't made a proper desktop-class chip yet. It's not because they don't care about higher core chips or chips with proper IO. They haven't made them because they CAN'T (yet).
M1 is still just a tablet SoC in a laptop form factor.
That wouldn't be good enough for production hardware. The developer units had a previous iPad’s A12 SoC in an enclosure for developers to test things on. For example, the A series doesn't support virtualization because that's not required on a phone or tablet. But the M1 supports this and other laptop/desktop features. Even the A14 only has 4 cores compared to the M1's 8 cores. This isn't just a tablet SoC.
We don't know what the efficiency will look like after you add all the I/O you need for a real desktop, we don't know how the memory subsystem will deal with 4 more cores that are just as fast. What's the point of having fast cores if you cannot feed them?
Perhaps you haven't been paying attention, but the Mac mini is a real desktop. And all of the M1 Macs have blazingly fast memory access. Remember, it's an SoC—the memory is on the same die as the CPUs, GPUs and they all have equal access to it. No bus to go across. That's why it's so fast.
People act like Apple changed the game or is light years ahead of everyone but they literally haven't made a proper desktop-class chip yet.
The M1 already has features no "desktop" chip has, such as 8 instruction decoders. This may not sound like a big deal, but none of the Intel or AMD chips--not the threadripper or the Zen3—has more than 4. Why this matters:
"It is because the ability to run fast depends
on how quickly you can fill up the ROB with
micro-ops and with how many. The more quickly
you fill it up and the larger it is the more
opportunities you are given to pick instructions
you can execute in parallel and thus improve performance.
Machine code instructions are chopped into micro-ops
by what we call an instruction decoder. If we have more
decoders we can chop up more instructions in parallel
and thus fill up the ROB faster.
And this is where we see the huge differences.
The biggest baddest Intel and AMD microprocessor
cores have 4 decoders, which means they can decode
4 instructions in parallel spitting out micro-ops.
But Apple has a crazy 8 decoders. Not only that
but the ROB is something like 3x larger. You can
basically hold 3x as many instructions. No other
mainstream chip maker has that many decoders in
their CPUs."
So Intel or AMD will just add 8 instruction decoders to increase their throughput, right? Nope:
"However on an x86 CPU the decoders have no clue
where the next instruction starts. It has to
actually analyze each instruction in order to see
how long it is.
The brute force way Intel and AMD deal with this
is by simply attempting to decode instructions at
every posssible starting points. That means we have
to deal with lots of wrong guesses and mistakes which
has to be discarded. This creates such a convoluted
and complicated decoder stage, that it is really
hard to add more decoders. But for Apple it is
trivial in comparison to keep adding more.
In fact adding more causes so many other problems
that 4 decoders according to AMD itself is basically
an upper limit for how far they can go.
*This is what allows the M1 Firestorm cores to
essentially process twice as many instructions as
AMD and Intel CPUs at the same clock frequency.*"
As I've said in other threads on HN, there's no way Apple's first laptop/desktop chip should be competitive with AMD, but it is:
"As far as I remember from performance benchmarks
the newest AMD CPU cores, the ones called Zen3
are slightly faster than Firestorm cores. But
here is the kicker, that only happens because the
Zen3 cores are clocked at 5 GHz. Firestorm cores
are clocked at 3.2 GHz. The Zen3 is just barely
squeezing past Firestorm despite hav...
Another example - M1s can't do SMT, no one is denying that they are good chips but there's a lot of catching up for them to do.
> And all of the M1 Macs have blazingly fast memory access. Remember, it's an SoC—the memory is on the same die as the CPUs, GPUs and they all have equal access to it. No bus to go across.
What does this even mean? Memory works at a certain voltage, clock speed, and has a capped rate at which it can transfer data if those are kept constant. For DDR4-3200 it's generally 3,200 MTs (megatransfers/sec).
What does every core has equal access to it mean? What does soldering SDRAM change in terms how many transfers/sec it can do? What do you mean there is no bus?
> The M1 already has features no "desktop" chip has, such as 8 instruction decoders.
What makes going wide a "desktop" class feature? Going wide isn't free, there is transistor cost associated. A M1 chip has 16 billion transistors for 8 cores. AMD's EPYC Rome has 39.54 billion transistors for 64 cores. That's 8 times as many cores. And that's including ~9 billion transistors just for the I/O die.
If you look at just the chiplets, AMD's 8 Core complexes take up only 3.9 billion transistors each. And these are 8 high performance cores with SMT, even removing GPU/NPU engines from M1, AMDs design clearly has a lower transistor cost.
> The Zen3 barely beats the M1
AMD was't trying to make an entry level laptop chip, so it's not good at being a entry level laptop chip? How do you think M1 would fare if we try to compare them on datacenter CPU metrics?
I understand why everyone is excited, but they aren't looking at the whole picture.
Just read "Why Is Apple’s M1 Chip So Fast?" [1] to answer most of these questions.
I’ll just hit the obviously erroneous issues here.
A M1 chip has 16 billion transistors for 8 cores. AMD's EPYC Rome has 39.54 billion transistors for 64 cores.
It doesn't make sense to compare the M1 with the EPYC Rome. They couldn't more different and were developed for dramatically different use cases.
(I also don't think it's not a good look when folks feel compelled to state Apple's M1 isn't as fast as AMD's $4000 64-core processor, as if this means anything. Of course the EPYC Rome is faster--duh. Anyways…)
16 billion transistors on the M1 SoC (emphasis on system) gets you an 8-core GPU, a 16-core Neural Processor, a Digital Signal Processor, a Secure Enclave, image processing unit, video encoder/decoder and of course 8 CPU cores and 8 or 16 Gb of memory.
Instruction decoders deal with the out of order execution of the instructions; more decoders enables you to have more instructions "in flight" as they say for the execution cores to run in parallel.
Again, you should read the article for the details, but due to the legacy architecture and technical debt of a 40 year-old instruction set, not to mention with instructions ranging in size from 1 to 15 bytes, it's not easy for the decoders to juggle all of the instructions to do out of order execution of instructions. Yes, even the EPYC Rome only has 4 instruction decoders.
Meanwhile, the instructions for Apple's Firestorm cores are all the same size, making decoding the instructions much easier and faster; so much so that the M1 has 8 decoders, which is unheard of in a mainstream processor.
Here's the bottom line quote from the article:
"In fact, adding more causes so many other problems that four decoders according to AMD itself is basically an upper limit for how far they can go.
This is what allows the M1 Firestorm cores to essentially process twice as many instructions as AMD and Intel CPUs at the same clock frequency."
Apple and AMD solved the "get as many instructions in the pipeline as possible and process them as quickly as possible" problem two different ways.
AMD does it with threading and many cores; Apple does it by increasing the number of instruction decoders and by having 60+ Gb/s bandwidth with a 128-bit bus. Apple didn't implement SMT because they don't need it; they have no issues with keeping their instruction pipeline full.
Bottom line: Apple's approach has more headroom than AMD's approach. They can't add more instruction decoders, so they crank up the clock speed and the threads to execute more instructions per clock cycle. But the M1 is so efficient, it processes twice as many instructions at the same clock frequency.
The M1 runs at 3.2 GHz; it's got a higher ceiling because Apple can add more cores and increase the clock speed a lot before it gets anywhere near the power consumption and other physical limits that AMD is starting to run up against.
> It doesn't make sense to compare the M1 with the EPYC Rome. They couldn't more different and were developed for dramatically different use cases.
The author of your beloved article compared M1 to Zen 3, and all Zen 3 CPUs use the same 8-core chiplet. From the cheapest 6-core Ryzen 3 5600 to the 64 core EPYC. All of them are made of the the exact same chiplets, just then number of chiplets varies. If you have 2 chiplet you get 16 cores, etc.
> Yes, even the EPYC Rome only has 4 instruction decoders.
It doesn't have "4 decoders"; it's just one 4-way instruction decoder. It's not the same thing.
> Firestorm cores to essentially process twice as many instructions as AMD and Intel CPUs at the same clock frequency.
A decoder doesn't not process instructions, it only interprets them and splits it up into smaller instructions. It doesn't make sense to compare number of instructions across different ISAs anyway, ARM has smaller, simpler, and on account of that more numerous instructions. One x86-64 instruction more often than not maps to multiple ARM instructions. It just makes no sense to compare counts.
I sort of stopped reading your comment around here, sorry about that.
Agreed - this is also why markets are inefficient and it can be easy to make money buying stocks.
M1 is a clear example of something a lot better where competitors are unable to keep up or adapt.
Yet tons of people in this thread still make up (bad) reasons why Intel/AMD have nothing to worry about.
I remember similarly dumb arguments about the iPhone too in 2007. Some people are just unable to recognize major change even when it's blindingly obvious.
Rosetta is a dynamic binary translator developed by Apple Inc. for macOS…
Follow the link for Dynamic Binary Translator[2]:
Dynamic binary translation (DBT) looks at a short sequence of code—typically on the order of a single basic block—then translates it and caches the resulting sequence. Code is only translated as it is discovered and when possible, and branch instructions are made to point to already translated and saved code (memoization).
Dynamic binary translation differs from simple emulation (eliminating the emulator's main read-decode-execute loop—a major performance bottleneck), paying for this by large overhead during translation time. This overhead is hopefully amortized as translated code sequences are executed multiple times.
Believe me, I know how Rosetta works. You don't need to quote Wikipedia at me ;)
Rosetta absolutely is an emulator, which has traditionally referred to taking code from one CPU architecture and making it run on another (in contrast with taking a binary of the same architecture but running it in another environment, which is often called a "compatibility layer" or in Apple's terms a "simulator", or the hardware-assisted version of this which is virtualization). There are many ways to write the part of an emulator that mimics another CPU; the simplest is an interpreter loop, but there's ahead-of-time and just-in-time code generation techniques as well. Dynamic binary translation is quite simple, as your quote mentions; Rosetta 2 instead achieves its performance through a combination of static binary translation to a form amenable to its runtime. Things that it cannot translate go through a fairly advanced optimizing JIT that goes way beyond what a dynarec would typically do.
Frankly I don't care that much about entry level macbook's absolute performance. But the utility of 20 hours 4k video play(3rd party tests even exceeded this number) can't be understated.
While it's not super relevant (for countries with high covid cases) now, not needing to worry about charging in the car, at the airport, on a plane -- those are real wins. Traveling for the weekend and forgetting your charger and... being fine.
Have you ever had your laptop on a train or bus and then found out the outlets aren't powered? It's stressful!
I'm not sure I see most of the disagreement here. As Gruber says, Moorhead is mostly complaining about software incompatibilities. But that's part of the total package--especially with Apple which is about selling the total experience. Personally I'll probably switch sooner rather than later especially given that I don't run a lot of native apps on my laptop. But it seems pretty reasonable to critique a new architecture laptop for (unsurprisingly) having some growing pains related to, especially, third-party software. Because people use a laptop to run software, not in isolation.
I had a suspicion upon reading Moorhead's review that he was trying to use familiar enterprise Windows software on the M1 instead of trying to use native apps:
> So far, I have experienced application crashes in Microsoft Edge, Outlook, WinZip and Logitech Camera Control. I got installation errors with Adobe Reader XI, Adobe Acrobat Reader DC, a Samsung SSD backup application, and Xbox 360 Controller for Mac.
What Mac user tries to use Edge and WinZip? A Samsung backup application? Xbox 360? Such a bizarre collection of software that snubs macOS's native browser.
To be fair, I'm getting absolutely enormous memory leaks in stock Safari on my new M1 MBP after a few hours of use. It doesn't seem to be reclaiming memory when tabs are closed.
There are growing pains, and they're not all in third-party apps. (It is hilarious to me, though, that WinZip a) exists for Mac and b) is used by anyone on a Mac.)
Not running mac currently, but one of the things I hated was how the UX was for thinks like keka etc, vs 7-zip/winzip interfaces... they just didn't match how I was used to using the things, also having to extract an entire archive to pluck out the one or two files I was looking for is often a pain.
I don't disagree re: the GUIs for Mac compression apps.
Part of the reason for this IMHO is because your typical developer on a Mac has always used the command line tools for that sort of thing. So, Mac developers were never really motivated to build great GUI tools for certain things.
Compared to Windows, where the built in command prompt was lacking for so many years, and as a consequence 7zip/WinZIP/WinRAR gained really robust GUIs.
It's ironic that you mention that, as I often used to just defer to the command line as the UX for the gui apps was so bad. I've always been centered around client/server or web app development or the backend and it was never painful enough to make something for myself. (scratch the itch)
I might be mistaken but, IIRC many browsers will not free memory from closed tabs. They'll keep it and reuse it for new tabs instead of deallocating and reallocating too often as you play with tabs.
Every new tab incurs more RAM usage; it's not reusing. Usage keeps growing and growing until the system gets sluggish and it prompts me to kill Safari.
A browser isn't viable to me without ad blocking and 1Password.
Firefox has issues currently on M1, as well - uBlock Origin won't load, for example. Thankfully, that particular issue is fixable by disabling javascript.options.wasm and javascript.options.wasm_trustedprincipals in about:config as a temporary solution.
I fully expect these to be fixed, but it's clear it's gonna be a few weeks/months before everything runs right. If I were a non-technical user, I'd be returning it for an Intel Macbook.
As far as I can tell, the issue is present without any extensions. I'm not willing to browse unprotected for several hours to 100% confirm, but it still grows with every tab.
Yes, it grows, but slightly. Ever since 1Blocker updated their app to use the older logo rather than the new monochrome logo I am not seeing the disastrous memory growth I was seeing before. After ~5 days of use without restarting Safari it is still hovering at around 800 Mb used, instead of the 10 Gb+ it was using before.
In the meantime, disabling the 1Blocker Button extension in Safari > Preferences > Extensions and restarting Safari fixes it. It won't affect how our app blocks intrusive content in Safari since 1Blocker Button only provides quick actions to the main app.
Ultimately, my point is not "I need help fixing this" - I've gotten Firefox working fine with a few about:config tweaks. My point is "non-technical users are going to run into weird issues they can't necessarily fix right now".
Wow, that’s a lot of memory use. Would you be willing to file a Feedback Assistant report with a sysdiagnose? Would like to see what’s up here. How many tabs is this?
I guess my question then is whether he uses macOS (note: it's OS 11 now!) as a daily driver or if he's just transplanting familiar software from his Windows habits in order to test drive the M1?
the chrome based edge is pretty decent... I'm still mostly using Chrome out of habbit, but there are a few things to like about the Chrome based Edge over Chrome.
chrome has enterprise managed mode just the same, with even more capabilities (i.e. spying features to keep your post wording) than the microsoft ones.
MS enterprise mode can only control settings and extensions, google goes several steps further.
That does sound a little sketchy... though I can see wanting to use the 360 controller, would like to see a review on how well most of the Steam supported games run for example. I'd been considering doing an ITX build for the living room, mostly for emulation/games... the new mac mini is definitely in the consideration, and seems to sit right between a lot of the nuc/sff options and mini-itx options and punch above it's weight against the latter.
Apple is pushing the "play any iOS game on the new Macs" hard. And a lot of them are supporting controllers natively on iPad and iPhone so obvious someone wants to try that out too.
Ditto Microsoft Edge, installed, ran a few websites - no issues, aside from a bit slow speed (although it may be just a crappy browser, never tried it on an Intel).
Perhaps with some heavier usage there are issues, but so far so good.
To be a "Mac user" you have to give up the software you prefer, and use the software Mac prefers you to use. And you don't get to be unhappy if the software you prefer doesn't run OK? I am so happy to not be a Mac user, even if I'm sad I won't be trying out the great M1 CPU any time soon.
Exactly. So you either go back to using macOS, or you complain that Windows doesn't run the software you like. It's OK to do that for any OS you run. It's a valid reason to not like an OS, or in this case, an OS that is going through a transition to a new ISA. Until it runs the software you prefer, it isn't ideal.
Sure, but it doesn't necessarily make it an objectively poorer product just because it's hard to swim against the current on that platform. Every platform necessarily has some expectation of doing things a certain way, and when things are done that way the experience will be better/easier.
You seemed to be implying that this was some unique problem with OSX and that's what is holding you back from trying M1. But really, it sounds like what is holding you back is just the normal barriers of changing platforms.
Personally as a Windows user I am hoping that Windows for ARM gets supported on M1 soon, but I don't think it's so unreasonable we're not there yet, and I am not going to discount the whole platform just because of it.
I do not hope to imply that this is a unique problem on OSX. I just disagree a bit that you have to conform to a standardized way to use a platform. Perhaps you do, with OSX, and it is a unique problem there? Obviously due to market structure, Microsoft gears Windows to "everyone" without being the "best" for many people, although it tends to be the best for PC gaming.
What is holding me back from trying the M1, personally, is cost. I just paid $950 for a Ryzen 7 4800H with 16GB and GTX 1660 Ti. It can't hold a candle to an M1 Macbook in battery life, but I know I can run my programs and play my games and not think twice about it. My failsafe is an electrical outlet. I'm hoping I can end up trying an M1 (or successor) through an employer though! I spent the past year using a ~2014 Macbook Pro through work, and for work things, it was better than my Windows machines for working on projects where everyone else used bash on a Mac. As WSL improved, I was able to be equally productive on my Windows machines and tend to use them and not bother with the Macbook Pro.
That's all a lot of extra anecdote you didn't need, but to be sure, I'm not trying to single out OSX - the OP singled it out by saying, essentially, "you're using it wrong!"
How is this different than any other OS? My preferred software is Mac or *nix based, so by your logic:
To be a "Windows user" you have to give up the software you prefer, and use the software Windows prefers you to use. And you don't get to be unhappy if the software you prefer doesn't run OK?
I can't tell but I think you're making my point. As a Windows user, I can absolutely be unhappy if software I prefer does not run. If someone comes around and says "but that software was originally for Mac... you aren't allowed to expect it to run well on Windows! A real Windows user wouldn't complain about that software."
And if the software I prefer runs better on macOS, I have a good reason to switch away from Windows. Just as if software you prefer runs better on Windows, you can complain about macOS (or in this case, incompatibility due to the ISA switch.)
My issue was with the comment using "mac user" as an excuse for software that isn't what the "true" mac user considers essential not working well.
MacOS is extremely good at supporting software originally developed on the OSes. After all you can get all the flagship MS Office apps on the Mac in very good iterations, plenty of Windows first vendors do support the Mac with version of their software, there are virtualisation and emulation solutions, and a lot of Linux software compiles and runs fine on the Mac as it's also a flavour of Unix.
In contrast, if you take the vast majority of MacOS first apps, there is no way you are ever running the vast majority of that stuff on other platforms. A wide variety of games even run on MacOS either natively or through compatibility layers like Porting Kit. In this respect out of Windows, Linux/Unix and MacOS, I think it's clear that MacOS wins hands down. It's not even close.
So given that this is pretty obvious, and I have a hard time believing you aren't perfectly well aware of this, why on earth are you trying to make this argument. Is running other platform's software really a primary criterion for you? In which case, the Mac should be right up your street.
Instead of a browser, I guess I installed demons. I wish they'd get this shit right. MS Teams always has this notification window hiding in the background so that whenever one alt-tabs to Teams the window doesn't actually have focus and your typing goes to /dev/null. And now their browser does this "fit-and-finish" crap that few muggles will parse.
Enterprise users often don't have a choice. Microsoft is pushing very hard to get Edge set as mandated browser at enterprises everwhere. In work we're also moving away from Chrome for this reason.
He may have wanted to say “the hardware is great, but not all software is ready for it, and it is consumer hardware, so it may not have enough ports for you”, but if so, it’s badly written.
Certainly, that second paragraph stating the M1 has many warts and is only fine for users who have a lot of money, to me, conveys a different message.
The disagreement is because most M1 reviews/comments were a bit of fan fiction - ranging from custom instructions for JavaScript, to "order of magnitude" perf per watt gains. Just as it doesn't make sense comparing the performance of the fanless M1 with a desktop CPU, it doesn't make sense comparing the perf-per-watt either; they are just optimized for different parameters. From what I've seen with the more objective benchmarks (not just geekbench), the Ryzens don't seem to be doing worse if you account for 7nm vs 5nm.
Docker not working sure was a deal breaker for me. That's the sort of thing I would expect to work out of the box. A huge number of developers use that daily and they seem like a pretty big chunk of Apple's target market for macbook pros.
I read Gruber frequently even though he annoys me (he's sort of like John Dvorak that way), and this piece is kind of emblematic of why he does. To be clear, the ARM Macs are way more exciting than I expected, and I'll probably buy one at some point. I also do not think Moorhead's Fortune article was perfect, but Gruber's righteous indignation is grating.
These statements are pitted against each other, even though they are both very much true:
> emulating or translating apps ... is necessarily going to be ... somewhat incompatible at best ... But Apple’s Rosetta 2 ... is a technical marvel.
He's upset about this quote from Moorhead's review:
> there were some very positive things about the new laptop. The new M1 processor is impressive, but far from perfect — it has many warts
Moorhead perhaps punctuated this poorly, but I think it's reasonable to assume that the warts refer to "the new laptop" in general, not just "the new M1 processor" in particular. Yes, his problems are generally software related (although he cited fan noise as a complaint), but that's a huge part of the product. Lots of people would love to use an M1 Mac without macOS, including one particular Finn.
Gruber can say, "There is no 'Well, here’s the downside' with regard to the state of Apple Silicon," but that remains to be seen. How much of the M1's advantage is due to a trillion dollar company buying up every last bit of TSMC's 5nm production capacity? Can they scale the unified memory architecture much beyond 16 GB? If they can, great; they've truly lapped the field.
> There is no balance, if by balance, you’re looking for a story that says any PC hardware, ARM or x86, is competitive in any way with the M1 Macs for low-energy computing.
This review wasn't about "low-energy computing." It was about using a new Mac as a replacement for an old Mac. This is a real, shipping product, not a beta. You can snicker at Moorhead's Microsoft-centric workflow and choice of apps., but that's his perspective.
>How much of the M1's advantage is due to a trillion dollar company buying up every last bit of TSMC's 5nm production capacity?
Yes, it's very hard to separate process from architectural (and implementation) design. And, as an industry, we're probably sticking our fingers in our ears more than we should about what happens when process shrinks really truly stop happening. And they will. There are still a variety of options but there's very little evidence they have anything like the legs of CMOS process scaling even if we do dump a big part of the problem on developers who don't get to develop for a largely unified architecture any longer or lean as heavily on high-level abstractions.
Complaining about incompatibilities in software used by less than 5% of Mac users and ignoring the strong compatibility of software used by 95% of Mac users seems misleading.
Equally, talking only about the stuff that works great and completely ignoring the incompatibilities that do exist is also misleading.
Every other review out there covered the strong compatibility and how amazing the M1 is. I for one am glad that there exist a least a couple of reviews that focus on everything that doesn't work. Let's be honest, no one these days makes a purchasing decision based off of just reading the first review they find.
I agree but these reviewers should weight and categorize the problems they find. For example reviews have been pretty consistent that
1) Graphic Design software:
Works well out of box. Key Adobe apps not yet native and have some minor issues under Rosetta, but still work, are still fast and should be much faster when native.
2) Video Production:
Unbelievably fast, on par with high end iMacs and Mac Pros and should get faster as all tools go native.
3) Software Developers:
Super fast builds and tools. Homebrew not native yet but can still use with Rosetta version of terminal. If you need Docker it’s months away.
4) niche Users of esoteric apps ported from windows: this is the group Patrick reported in. Yea, some of these apps are months away from being usable but all have perfectly cromulent native replacements.
The laptops have a lot more surface to dissipate heat than a phone, and less likely to operate wrapped in an insulating case. Also, that some models have active cooling can make a difference here.
A vast understatement, IMO. I've gotten my iPad pro hot enough you couldn't hold it in your hands (well, I couldn't) while running graphics intensive hardware.
IDK how it actually went down at Apple HQ but I can imagine. They are so, so keen on sleek, quiet machines that it tops all other concerns. When the hardware engineers said "more performance is going to require a fan or two" the leadership said "nah, let's design and fab our own CPUs".
More likely that they realized their iPad chips were faster than their Intel notebooks, and really had no choice but to move to their own chips on the Mac.
Joke with a kernel of truth. I'm not saying they didn't do a great job with the M1. I'm just saying it's an unusual direction for their business to take unless it was initiated to serve their core strategy of delivering visually appealing devices with nice software.
That's my impression too. The previous post currently says 11 hours ago, so by my count that's midnight NYC, 5am London. I know Australia exists, but overall that just gives it most the night to slide down the page before the majority wake up.
Both can happen at the same time if submissions within a window of 24h count as points to the original one.
Try to submit this same link now and that's what will happen. This feature already exists, it's just that it seems too narrow of a window if it's only a few hours long.
To me, this is a community of people that want to share interesting topics and information, and contribute to that sharing and discussion. The greatest prize of all this is how much accurate and useful information we all get out of this. I've become a veritable encyclopedia, a proficient and effective software engineer, a resource on many topics for friends and family, and continue to be an enthusiast of computing software and hardware advances.
There's no denying the part of our brain that gets a kick out of the internet points which reflect peer approval and respect. However, I will continue to hope to value the greater good that this community creates for all of us over the individual recognition that gives me a much shorter term buzz.
Wholeheartedly agree and couldn’t have said it better myself.
Aside from the peer approval that you mentioned, I also enjoy the idea that I can somewhat steer the conversation and see what smart people have to say on the topics I’m most interested in.
Yes, sorry—I know it's frustrating when a later submission 'wins'.
The reason we allow reposts of articles that haven't had significant attention yet is that it's largely random what happens to get noticed on /newest, and we want to allow good articles multiple rolls of the dice. The priority is on the article, not who submitted it. That does make it a bit of a lottery, and it's winner-take-all. There are definitely better solutions possible, and it's on my list to work on that someday. We'll expand how karma is reward for submitting an article (HN could use some karma inflation anyhow) and find some way of grouping related submissions together.
In the meantime, the lottery does at least even out in the long run if you keep posting good submissions.
Are people really harvesting links on /newest? You guys should know better of course, but I always thought what drove an article to the top was multiple people submitting the same link. That was my reasoning for a 24h window defining a submission.
As for karma inflation, it's an interesting concept, but I don't know if I'd like that :)
Sorry, I meant that I for one don’t go to /newest often.
And I didn’t imagine it was what drove most links to the top of the homepage. I always thought that many people submitting the same URLs was more influential.
I don't care about any Apple products, but I really do hope that M1 will force all other manufacturers ship good ARM laptops. Or force Intel and AMD to pull a rabit out of their hat to match the performance/watt of ARM chips.
I want long battery life, no fans and a lot of memory. Running Linux. If M1 is the catalyst to make it happen, then so be it.
I agree. Competition is awesome for consumers. I don't care much about macs either (although I'm really interested about Mac Pro Mini rumours), but I know that Intel and AMD will invest lots of money into research right now as not to become obsolete.
Don't get your hopes up. My prediction is, that yes, there will be a slew of new ARM laptop's. And even though the technically run ARM chips, it will only be a marketing selling point for new. They might even cram Snapdragon's or other mobile device CPU's in there, just to be able to put a ARM sticker on a plastic frame with an folding LCD panel and keyboard. They will sell it as beter than Intel, because that is what people who can afford an Apple will tell them is so great about their new laptop. In the long run, due to this, the Windows ecosystem might adapt to embrace ARM as a whole and then we will start to see competetive ARM PC's. But it will take a while.
Ultimately, it's all about Windows. I know it's hard for many of us who basically hardly see anything other than Mac or Linux laptops to really grok, but something like 90% of the laptops out there run Windows (if you exclude Chromebooks which AFAIK are mostly used in education).
Indeed, this is why Apple is able to take this leap and the PC platform hasn't so far. They don't have to wait for as many partners to get along for the ride.
And partners also means customers. A ton of those Windows systems (and I should have said systems--rather than laptops--as I don't know how things split out on laptops specifically) are going to non-technical users in big companies who absolutely do not want any changes that alter a menu or move an icon or introduce some incompatibility they have to deal with. They absolutely do not care whether there's some significant price/performance/watt boost if it otherwise introduces even mildly disruptive change.
Frankly, I think that the willingness on the part of some, like Microsoft, to cater to that degree of hidebound thinking is actively harmful to us as a society, especially in terms of technology.
The more we enable people who are unwilling to even try to learn that this particular option has moved from the Edit menu to the Format menu, or has removed the default toolbar icon (but you can add one), or has simply moved an inch to the left, the more we send the message that it's normal and expected that people shouldn't have to ever adapt and change and learn new things.
On the other hand, we have a lot software in the industry that changes regularly without any perceivable improvement. Google is known for relentlessly pruning whole products that people were happily using but they're also pretty aggressive about removing features from existing products.
One big example of that is the removal of community captioning for YouTube videos. Call me hidebound but I am not happy at all with this, even though I don't use it very much. Captions are critical for making videos accessible to the deaf and hard of hearing communities. The removal of this feature is an ugly snub to a vulnerable group. Now they will be wholly reliant on the video creators to do the captioning themselves or otherwise be stuck with the lousy automatic captioning.
I'm not sure there's another company of Apple's size and influence which has consistently been so ready to dump legacy at least a beat before its customers were prepared for it. Headphone jack, 5 1/4" floppy, I'm sure a bunch of other things I'm forgetting.
"I'm going to dump Apple over this!"
"See ya. Don't let the door hit you on the way out!."
Does anyone miss floppies? CDs? What about SCSI or ADB? I can't think of one thing Apple dropped that caused such an uproar at the time that people regret now. Sure, I bitch about my laptop not having an SD reader, but Apple hasn't shown me what's next for its replacement other than thinner (meh). SCSI->USB/Firewire->USB2->USB3->USB-C this all made sense.
You're right; I was accustomed over the years to point to PS/2 as an example of the PC world being slow to give up on an outdated port, and carelessly conflated that with Apple using it.
It's also because Microsoft half-assed their Windows-on-arm attempt. They made something akin to Rosetta 2 (x86 to arm translation) but only supported x86, not amd64. Apparently they are working on 64bit support, but releasing something half done doesn't inspire confidence. Even the 32bit support was reportedly not very good/fast, compared to Rosetta 2, anyways. If Microsoft steps up their game then I can see Windows on arm starting to take off, even if Qualcomms chips trail Apple's M1 in performance.
What makes it mobile, the device it's in? If I have one in my desktop or Mac Mini, it's no longer mobile. The power draw? Everyone benefits from it using less power regardless of use. Form factor? Why does size physical size dictate small == mobile?
I was responding to a post that talked about putting Snapdragon or similar ARM chips in laptops in a deprecating tone, and was making this point exactly.
> Or force Intel and AMD to pull a rabit out of their hat to match the performance/watt of ARM chips
I believe I read an article the other day that explained why the M1 was so fast and efficient and basically the design it came up with is impossible to do on Intel, for many reasons including the fact that Intel instructions are variable-length (which means they need to be completely parsed every time to find the end of the instruction) while M1 instructions are fixed-length (which means you don't need to parse it) due to the RISC design (finally a winner).
Methinks we are witnessing the death of the traditional Intel chip API as the bad design decisions made along the way come to a head.
Should make things interesting as now both Apple and Linux have a head start on ARM support.
Apple had versions of iOS running ARM on phones and pads since how many years? They used their head start already to launch themselves into their ARM computers.
ARM was a joint project between Apple, Acorn and VLSI, so you could say that they had a head-start on pretty much everyone except Acorn, who invented it in the first place.
Most UWP apps should be easy to compile for ARM. Emulating x86-32 apps on ARM Windows has been a thing for many years and the further expanded x86-64 support is in Insider builds.
Really the bigger issue is that the ARM chips Windows has thus far run on don't have the similar TSO opts the M1 came with.
Funny thing, you could make a constant length superset of x86-64 with backwards compatibility, and it's even going to run fast...
(The instruction decoder would become more of a translator.)
The bigger problem with x86 is the implicit hard synchronization barrier after most writes. This cannot be easily fixed without introducing new instructions altogether.
> Or force Intel and AMD to pull a rabit out of their hat to match the performance/watt of ARM chips.
AMD has been great (and improving) on performance/watt. It appears that the M1 has the upper hand, but then again Apple bought all the 5nm capacity for the year. I'm really curious how well 5nm Zen chips would compete non performance/watt - I suspect Zen 4 (5nm) will compete very favorably.
More ARM SoCs means more machines that will never see Linux or *BSD support. A lot more work[1] needs to go into supporting each individual SoCs than x86 machines require.
I have a four year old MacBook that I am still happy with, but when LispWorks supports M1, I think that I will get the small M1 MacBook Pro - I can live with an extra pound of weight.
I use my iPad Pro for just about everything except programming, so laptop selection is not as important to me as it used to be.
I think the missing part of the discussion is that the M1 can run fast and hot... It hasn't transcended physics - it's just not being pushed hard enough.
Give it a 800% overclock and dunk it in liquid nitrogen and then we'll talk.
It will conk out at rather low clocks as the design is not using as many high frequency tricks such as local clocks, instead going for high efficiency.
No device this is going in has a large TDP budget. A unit tuned for a desktop heat sink will put in some interesting numbers. I'm curious to see when they put out a Mac Pro with an M-series chip. Or if server is ever coming back.
The distinction here is between "fast" and "fast_er_". It isn't that it can't run faster and hot (because of course it could) .. but that it's able to achieve fast performance relative to the competition without having to get "hot".
I feel like the M1 hype is getting out of hand. Apple bought up TSMC’s entire 5nm manufacturing capacity so of course it’s going to pack more transistors and use less power than AMD’s 7nm TSMC process or Intel’s 10nm process (which is effectively similar to 7nm TSMC).
The M1 is a great and well-refined design but when compared to AMD’s similarly priced 4750U in multi core benchmarks, the performance per watt is better but not by much more than the improved process would suggest. And that’s without the IPC improvements that mobile Zen 3 will bring.
And when compared to last gen 14nm Intel MacBooks, a 5nm TSMC part is going to blow the doors off just by virtue of feature size alone.
It very well may be over hyped but with the breakdown of Dennard scaling[1] the gains we've seen in CPUs has been pretty minor generation to generation.
I think that could also be the reason it is getting so hyped, it is a tangible improvement over other CPUs. 50% more battery life, better performance in some case, what's the last CPU generation we saw such gains? Probably pretty far back to when Dennard sclaing was still a thing?
We have seen massive improvements in power envelope in CPUs in fairly recent generations, to say nothing of the power improvements seen in some generation->generation GPU improvements.
For example, with an improvement to the 14nm process and not a die shrink, Kaby Lake->Whiskey Lake saw low power processors double in core counts (2C/4T->4C/8T) in essentially the same power/thermal envelope. Basically every thin laptop/ultrabook family doubled its core count.
Well... yeah... but they also dropped clock rate from 2.5GHz-ish to 1.7GHz-ish. That could equally well explain the increase in core count at the same TDP. You're gaining about 15% IPC improvement from Kaby->Whiskey [1].
It's an overall improvement, but not as dramatic as "2x cores for 2x perf at the same power"
"Base" clock dropped, but boost clocks remained fairly high. In practice performance gains were quite good (except, notably, for that time when Apple used old power control firmware and had 6C/12T processors underperform their 4C/8T predecessors).
The point of this is that the significant improvement from Kaby Lake to Whiskey Lake involved only small architectural refinement and updates to an existing process, so much larger performance/power improvements should absolutely be expected from an entirely new process plus architecture refinement.
I'm on a 3 month wait list right now to get an 8 core/16 thread AMD Thinkpad. Nobody else currently produces a sub-15W part with that much computational power.
The major difference vs the M1 is that AMD's micro-PCB design scales in core count rather than producing a single dense power-efficient die.
> I'm on a 3 month wait list right now to get an 8 core/16 thread AMD Thinkpad. Nobody else currently produces a sub-15W part with that much computational power.
This is a big part of the reason that the AMD mobile chip comparisons seem to be kinda besides the point for me. For a bunch of reasons, its almost impossible to get a high quality laptop today w/ the latest and greatest AMD chips that people love to compare to the M1 chips. I know AMD makes great chips, but what does that matter if you can't get buy them in a laptop today?
FWIW - I love AMD, I have a Ryzen desktop for gaming. Its fantastic the competition they are bringing in the desktop market. But I don't want to play the game of searching far and wide for high quality laptops that have the new Ryzen chips _and_ are actually in stock somewhere in a configuration I want _and_ are from a reputable manufacturer.
I was able to pickup a M1 Air on release date w/ the upgraded specs from my local Apple store. It's a glorious machine, and I knew the build quality and hardware is going to be top notch.
OEMs still largely prefer Intel for laptops for whatever reason, which sucks, and it seriously hampers the ability for AMD to compete in this market.
>OEMs still largely prefer Intel for laptops for whatever reason, which sucks, and it seriously hampers the ability for AMD to compete in this market.
OEMs might prefer Intel because, unlike AMD, Intel is actually able to supply the CPUs to go in their notebooks. A slow CPU is better than no CPU. My understanding is that AMD is mainly focused on supplying the gaming console makers, so I don't know that the situation will improve over the next few years for laptop manufactures seeking AMD chips.
Sure, as I said, I think there are many reasons that AMD has largely not been able to make a dent in the laptop market.
I do find it amusing the continual retorts to M1 performance of "but wait until the Ryzen 4800/4900 laptops are out in force!" as if that were something that is realistic at all in the next year. AMD can't even keep up w/ demand for its graphics cards and console chips right now, unfortunately.
> It was a fundamental trade-off inherent to PC computing, and now we don’t have to make it
I wonder how many of the currently accepted trade-offs are actually false.
At low levels it's kind of easy to prove, like you can store either more or bigger packages in a truck but not both. But the higher you go, the farther from physics, the more difficult it becomes to know if you're making decisions on what is essentially a flaky trade-off.
It turns out that the benchmarks for M1 vs latest generation Intel & AMD CPU's are indeed overblown, and it is an incremental improvement more than a great leap forward.
The source of the confusion has been the benchmarking software. To saturate one core on an Intel processor you need to run two threads, because that's the way they are designed. So the single thread benchmarks that have been used so far have been using 50% of the capacity of an Intel CPU core and comparing it with 100% of the capacity of an M1 CPU core.
I didn't until you linked to it, but having read it it still does appear that comparing single thread to single thread is not very accurate. Neither would core to core considering single thread does hold some weight compared to real world workloads.
I guess at the very least CPU benchmarking software should have a "thread to thread" benchmarker alongside a "core to core" benchmarker, or something along those lines.
That would be in the spirit of having benchmarks indicate real world usage
Of course, the scheduler can easily be told to run both benchmark threads on the two hyperthreads of a given core.
The argument to be made is that because of the resources dedicated to SMT, single thread on AMD/Intel versus single thread on Apple is not measuring the true potential performance of the whole core. In principle, some multithreaded workload over all available threads could be a better metric for whole-processor performance.
Fair enough.
But if you're comparing single threaded performance, there isn't a reason to split the workload into two threads for AMD/Intel and have it as a single thread for Apple.
If I had a single threaded application or a single threaded critical path of a multithreaded application.
Benchmarking multicore Performance+Efficiency (Apple) versus SMT (AMD) versus SMT+Wide vectors (Intel) is never going to provide perfect apples-to-apples comparisons. There's an entirely reasonable argument that single-thread performance is oversold as a metric, and that the focus on it advantages some platforms over others.
At the end of the day, benchmarks are inherently only an approximate measure of how real-world code will perform. SMT, basically by definition, is rarely going to benchmark well, but is inherently going to show more of a benefit when running real-world mixed workloads.
Eh? I mean, by the same token you could say that single-core benchmarks of Intel chips are invalid, because you need eight threads or whatever to saturate a POWER7's multi-way SMT.
The main purpose of single-threaded benchmarks is to approximate performance for things which are actually single-threaded.
I've never owned a Mac and don't intend to own one but I'm very happy with this M1 move, mainly for the reasons highlighted in the introduction to the article:
>M1 Macs embarrass all other PCs — all Intel-based Macs, including automobile-priced Mac Pros, and every single machine running Windows or Linux. Those machines are just standing around in their underwear now because the M1 stole all their pants.
I think it's a bit unfair to lump Linux in there since it's been running on ARM and other embedded, low-power devices basically forever, but it is true that few of us run our Linux desktop on an ARM board.
But I think the overarching point is true: now desktop and laptop makers can't just pretend that having a hot and/or slow, clunky, noisy x86-based architecture is just a fact of life. I hope and expect that it'll help create a new generation of ARM-based laptops and maybe even desktops that will run cool and smooth.
Add 50 more PCIe lanes, double the bus width, 270mb of cache, 60 cores and a communication protocol then get back to me.
Anandtech ran the numbers a couple years ago and an EPYC processor spent around 60% of its budget on Infinity Fabric (up to 90% of used power with only a couple cores being used).
An M2 based chip might get more done, but it won't use much less power. In truth, I'd guess there's a good chance it bottlenecks on the IO and performance doesn't increase as expected (without widening IO further with all the extra power costs).
I really can't wait to see what they'll do in the desktop space. For all the gushing reviews the M1 is receiving, it's worth remembering this is their entry-level machines.
As it stands right now, I have an 18 month old 27" iMac with 64GB ram and an AMD gpu - and an 8GB M1. And I'm preferring the M1 for almost any task that doesn't demand the real estate.
>Wait for an 8-core (or the rumored 12 core) variants. Those will embarrass your threadripper.
It will not, my TR has 64 cores 128 Thread and you know whats the best thing? 128GB Ram...and imagine that..they aren't even soldered...and i even can exchange my GPU..or slap in 2 others.
And if apple ever brings out a M1 Workstation, i can probably buy 3 other Server/Workstations with that money.
Your TR is what, 280 watts TDP? nearly 20 times more than a M1. And is 30% slower in single core.
Firestorm Cores use less than 4 watts each. A 16 firestorm core Mx will produce roughly the same multicore performance as your TR, with a less than 70 watt TDP.
And a $65 M1 dusts a 4800h that costs over 4x as much and using 1/3 the power. Top end ThreadRippers cost $2,000 to $3,000 each. You can bet an Apple Mx ThreadRipper killer isn’t going to cost more than $500 given it only needs 4x the cores of the $65 M1.
It matters because its not just adding electricity, its adding heat. Pretty clearly Apple can add a whole bunch of Firestorm cores to future M1 derivatives and still fit in smaller power/heat budgets that current generation Ryzen and Threadrippers operate in.
Maybe. I don't know enough about CPU design to argue with that. There are others elsewhere in this thread saying that a lot of the energy and heat goes on the communications systems and other resources you need more of when you ramp up the core count.
I suspect we'll never know. Apple have shown little interest in anything other than prosumer workloads for a long time. I doubt they'll be making an M1 server platform anytime soon.
I have to somewhat concede your last point. What percentage of AMD sales are Threadrippers in any year?
It's possible Apple just doesn't think the investment in a 30 core CPU is worthwhile and stops around 16 cores, which would still be higher than any i7 or any i9 ever made.
Given that scaling, you're looking at 140w between 64 cores which is right at 2w per core. The all-core turbo for TR is 3GHz. If we assume linear scaling on M1 cores, if you dropped them in to replace the x86 cores and the had to reduce from 3.2GHz at 4w down to 2w per core, the all-core turbo would be a mere 1.6GHz AND about a third less cache per core.
TSMC claims 30% less power going from 7nm to 5nm. That means those x86 cores power usage drops from 2w to 1.4w. With the same linear scaling and 4w at 3.2GHz, the M1 design is now operating at 1.1GHz competing with a part that is still 3GHz and the M1 still has only 3mb of cache per core instead of 4.5mb.
Now, I'm betting that actual power usage is 3w which bumps frequency up to 1.5GHz. Still half the frequency, but possibly close enough in performance to Zen 3 at that frequency.
EDIT: It's apparently 3.45w during cinebench according to these benchmarks.
I'm betting that if they didn't bottleneck on the IO that the M1 would perhaps be faster, but at that scale, interconnects are a much, much larger issue than the cores themselves.
Zen 3 chiplet is 4.15B transistors. The consumer IO die is 2.09B transistors. The Threadripper IO dies is 8.34B transistors. M1 is 16B transistors. A14 is 11.6B transistors. That difference is mostly 2 firestorm cores/cache, 4 GPU cores, another DDR4 controller, and some Thunderbolt/PCIe.
Taking a ruler to a die shot and multiplying that area by transistor density (16B/119mm^2) gives us somewhere around 500M transistors per core plus 3mb of cache. Adjusting out to 4.5MB per core to match Zen 3 adds another 75-100M transistors.
Zen 3 gives 518M transistors per core, but that includes interconnects. Taking a ruler to the core reveals about 14% of the die is interconnect, SMU, and wafer tests (about 580M transistors). Subtracting that from 4.15B and dividing into even core/cache gives 446M transistors per core.
We'll assume an identical IO cost of 8.5B transistors and 150M transistors of interconnect per core. (64 * (600 + 150)) + 8500
is 56.5B transistors and 420mm^2 total area. A zen 3 chip with similar densities would be 310mm^2.
No matter how you slice and dice that, Apple's chip is going to be significantly more expensive than AMD's chip. Thinking anything different is a pipe dream.
EDIT: also note that formal verification adds a lot to chip pricing.
This is just not true. A mini going from idle to a full single-core load is a power increase of 6W. A heavy multicore load hits over 25W.
> A 16 firestorm core Mx will produce roughly the same multicore performance as your TR
With what fantasy math? 30% faster single core performance would be 16 * 1.3 = ~21 equivalent TR cores. It'd be less than half the performance of a 64 core ThreadRipper.
> And a $65 M1 dusts a 4800h that costs over 4x as much and using 1/3 the power
On-die memory is pretty much their one ace in the hole that nobody else will have the guts to match. Intel can't because their fabs are behind, and AMD probably doesn't have the resources to fork their architectures and/or piss off the server integrators.
I can't live without 64GB in my desktop and would love to have more, so the new Macs are not for me.
M1 isn't using on-die memory[1]. It's on-package, and likely little more than just off-the-shelf Micron or Samsung dram modules that are just soldered instead of socketed.
It's a pretty accurate cost estimate from those tracking Ax series costs. A chip AMD sells for $300 doesn't cost AMD $300 to make, but being a public company their gross margins tell us it costs roughly $150.
But you were comparing this theoretical production cost to the retail cost of other chips, that doesn't make sense.
Hardware companies don't charge cost + some reasonable markup, they charge whatever the market is willing to pay for it. I highly doubt Apple would release some hypothetical "threadripper killer" at a price any lower than they needed to to compete.
My point is that someone building a Threadripper PC has to pay $2,000 to $3,000 just for the CPU.
Lets assume that all the other parts, high speed RAM, SSD, power supplies, cooling, motherboard, etc cost another $500, and that operations, assembly, marketing etc average 20% and the manufacturer aims for a 5% margin. So they sell their low end Threadripper PC for $3300, and their high end Threadripper for $4,700.
Then Apple comes along and makes it's own high core count versions of the M1 (lets call them the M16 and M24) for $1,000 and $1,500 with similar performance to the Threadrippers. When Apple builds high end Mac Pros out of the, they also spend $500 for other parts, giving total parts cost of $1,500 and $2,000, it marks them up 50% for their typical margins, and sell them for $3,000 and $4,000 and have the same performance as the ThreadRippers from the lowest margin manufacturers at for 10-15% less.
And on top of that, the M16 and M24 use less power and generate quite a bit less heat than the Threadrippers. They don't need any fancy cooling systems. That can lead to smaller packaging and even more cost savings.
Obviously when Apples 16 and 24 Firestorm core Mx chips come out newer Threadrippers will have evolved to smaller process as well, and that will help increase their performance, while reducing cost and heat. But there is little doubt that Apple's Mac Pros will be much more competitive price and performance-wise at that time, and if they keep leveraging TSMC first there is a significant risk that they actually take some of their laptop cost/performance advantage up to Threadripper levels.
> And if apple ever brings out a M1 Workstation, i can probably buy 3 other Server/Workstations with that money.
Based on the MacBook Air/ Mac mini/ MacBook Pro pricing, I don't think this is the case.
Apple has delivered a significantly faster MacBook Air for the same price as the previous generation. Not only did they increase the CPU speed, they essentially doubled the performance of their the SSD while maintaining their base price.
Apple dropped the price of the base mini by $100 while delivering a significantly faster Mac mini.
I suspect whatever Apple delivers for their top end machines will be equal or lower price to their current Intel machines.
He's talking about the vast majority of PC chips sold today, who's power consumption / performance / thermal ratios just got marmalised.
You're quite right, your machine and in fact everyone else's are still just as good as they were a few weeks ago. I got an Intel 5K iMac in the summer and I'm still very happy with it. I need to be able to run windows and Linux in VMs so M1 isn't for me right now. It's no for you either.
Having said that, your manufacturer of your Threadripper is now on notice. Apple isn't done with the M1, more powerful desktop chips are on their way over the next few years and there's a very good chance they'll do exactly what they just did to Intel and AMDs mobile chipsets to the high end stuff. Apple are coming for them, and not being impressed now isn't going to serve them well down the line.
It's all going to come down to user experience and versatility. Apple is doing some crazy stuff with the M1, but I personally would chock (most of it) it up more to being clever than true innovation. I think only Intel needs to be concerned right now because they're being outpaced by both AMD and Apple. Additionally if the acquisition of Arm by Nvidia results in what people expect, Intel will basically be losing out in every market.
Going back to versatility, there's no way that AWS or Azure is going to move to Apple silicon for their infrastructure, so there's still a massive market for AMD or other Arm manufacturers to target. For the consumer side unless Apple releases a chip that results in them making a sub-$800 laptop, they're just simply not going to capture a large market. With the rise of devices like Chromebooks in education and home there's still something to be said for Apple devices just being too expensive.
In general, Apple devices also leave a LOT to be desired for any sort of large scale enterprise organization. Unless they go back to their 90's and early 00's mindset, I just can't see this changing.
It's certainly exciting to see the CPU market being upended by Apple and AMD, as well as the potential future for a large architecture shift, but I don't think there's any real threat to market shares as a whole.
You don’t just go and make ever part of your chip architecture wider by being clever. Well, there was probably some cleverness involved, but it’s not like this has been a one-off trick.
By clever I mostly mean, from what I've seen from reports, handling stuff like transition from x86 to arm by baking in instruction translation into the die. Their approach with some things is a bit different from what others are doing, but it's not like they wholly invented a new 2nm process that breaks what we thought the laws of physics were, they're using the same fab as everybody else. Thus their design at the moment is clever.
there is little hardware assist though: TSO can be enabled and disabled and thus simplifying x86 instruction translation by implementing the x86 memory model in hardware and thus avoiding unnecessary memory barriers
Right, but they're going to remain a tiny niche. They're priced so highly that there's no risk of them cannibalizing Amazon's own AArch64 instance-types.
I don’t know about under $800, but I wouldn’t be at all surprised if Apple didn’t keep the current M1 Air on the
lineup at a lower price when the Airs get their next revamp.
This is exactly what Apple did with the iPad, nowadays you can get a very decently specced 10.2” iPad with their 2018 CPU for $309. That’s astonishing, but it’s a great way for Apple to get extended value out of their silicon.
Sure, and when they do, we can have that conversation.
But claiming that what’s available today from Apple is of great interest to the high performance market is jumping the gun.
There’s also the fact that cost (including peripherals and support) and software availability, not performance and efficiency, are what guide most PC purchases. The M1’s advantages are far less pronounced in those areas.
> But claiming that what’s available today from Apple is of great interest to the high performance market is jumping the gun.
I think that misses the point that most users (and most software developers in particular) use laptops. The M1 SOC pretty well kills everything CPU-wise, and does graphics as well as a GT 1050 discrete GPU. It also throws in AI silicon and RAM, and runs all of it in a 20W (or so) power budget.
Single-thread performance is quite comparable to a Threadripper.
It'll be interesting to see how things look when M1 style silicon is introduced for the 16" Macbook Pro (guessing 8 performance cores, 32 GB RAM), and the Mac Pro (guessing up to 32 performance cores, and at least 64 GB RAM used as cache for a potentially much larger main memory).
> There’s also the fact that cost (including peripherals and support) and software availability, not performance and efficiency, are what guide most PC purchases. The M1’s advantages are far less pronounced in those areas.
Actually as to cost, it's highly competitive in the premium laptop segment (like Dell XPS for instance).
Software availability is good, since it'll run almost all Intel Mac software, all M1 Mac software (soon to be ~everything), and all ARM-compatible Linux software (ARM Linux runs in a VM).
The only thing missing is Windows, and that may well be temporary.
I'd say if you're doing web, Mac or Linux software development the Apple Silicon Macs are the wave of the future! :-)
I'm the other side. Vertical integration is a fast way to fall behind.
This move is very Nintendo-like. Gimmick then don't even attempt to compete with the big dogs.
The fanboys have their meat to chew on "it's not hot". And they will repeat it on command as Apple falls behind to a swath of competition. "It might be slow, but at least it doesn't warm my lap."
Nintendos profit forecast for FY2020 is 150% of Sony's gaming division.
Apple just dropped hardware which crushes every comparable competitor.
But to you, Nintendo is not competing by "making more money", and Apple is "falling behind" by ... jumping years ahead? This reads like personal axe-grinding that is detached from any objective link to the market.
The third most valuable company on the planet's long term business model has been a "fast way to fall behind", and is not "even attempting to compete with the big dogs"?
Well, they just pantsed Microsoft at their own ARM laptop game (#2 most valuable) and the #1 big dog is, er, Saudi Aramco.
Before the M1 I worried about that. Now I don’t. The M1 is the lowest performance, lowest end Mac chip Apple will ever make, and it’s already knocking on the door of premium Intel and AMD chips. Not the super-premium, highly multicore stuff like threadripper, buts it’s a lot closer than anyone expected in single core.
To me, this means that when Apple consciously goes after the high end desktop market, they are likely to overshoot in exactly the same way M1 overshot the mid range mobile space. That could put Apple’s high end iMac chips in Mac Pro territory anyway.
but my threadripper's are not impressed with some 4 core M1
The M1 has 8 CPU cores, 8 GPU cores and 16 Neural Engine cores. It also runs at 10 watts and maxes out at 24 watts. A Threadripper starts at over 100 watts…
To be fair, 4 of the cores on the M1 are efficiency cores and are very low performance (but sufficient performance, at incredible efficiency, for a lot of casual ongoing background things). An 8 performance core M1 would be a monster.
Not sure if it maxes out at 24 watts. Someone measured wall power of a Mac Mini hitting 21 watts at full load, but that's including power supply inefficiency, all of the supporting circuitry and IO, a fan, flash storage, etc.
> 4 of the cores on the M1 are efficiency cores and are very low performance (but sufficient performance, at incredible efficiency, for a lot of casual ongoing background things).
I think you're underselling the small cores a bit. Their microarchitecture is roughly comparable to a Cortex-A75, which was the high-performance cores in Snapdragon 845 phones like the Samsung Galaxy S9. But Apple clocks them about 26% slower than the SD845 did, so maybe we should be comparing against Galaxy S8 instead: flagship Android performance from 3 years ago. That's capable of a lot more than just casual background processing. Samsung introduced their DeX docking station for smartphone-powered desktop usage back with the S8 generation, when their phones only had as much computational power as the Apple M1's smaller cores.
Pulling other cores into this discussion does nothing but muddy the waters.
The Firestorm cores in the M1 are 4.5X faster at SPECfp than the Icestorm cores. The Firestorm cores in the M1 are 3.1X faster at SPECint than the Icestorm cores.
In my world, the 3.1 - 4.5X faster sibling makes them very low performance, relatively. In a fully working together scenario, the "8" core M1 is equal in performance to a conceptual 4.5 core Firestorm part (see the Geekbench scaling, for instance). This, again, has positively nothing to do with any other core or part.
Bizarre that anyone thinks this is somehow negative or dismissive of the M1. It is an extraordinary chip. But they intentionally, purposefully put 4 very low performance -- compared to Firestorm -- cores because it matches the computing model of users average workload.
No probably not. More of a "Benchmark vs. Power Consumption" war I think.
Power users understand today that max MHZ doesn't even matter anymore, for various reasons.
I think power consumption (even in desktops, servers) will be a large factor / marketing number.
I think the big tech sites will continue to hammer home some type of "CPU work / watt" type metric, which is roughly like "time to complete some standardized workload vs. power used to do it"
No, but we've had performance per watt wars for years now. It's why prior to Zen you only saw AMD chips on the cheapest of laptops, and also now why Intel is panicking.
People buying a threadripper today don't care about it.
What matters to them it's the performance, my gf is a 3d artist, she doesn't care if some cpu consumes more or less power than another (when I say she doesn't care I mean she doesn't even know that CPU have very different power requirements), she needs the fastest gear she can buy on her budget.
Apple M1s are not it.
There are many other that do not care because the hardware stays at the office, where someone else pays the bills and they too don't care because the electricity bill is the last of their problems, fixing it buying more efficient computing devices would mean spending a lot of money in advance to replace at least half of the stock.
How many more months of electricity could that money pay?
A lot.
People buying Apple for its M1 low power consumption are a niche inside a niche.
So I think the reasoning stands: Apple M1s are not a real threat for its competitors because the market that really cares about its strengths is smaller than the one that doesn't and Apple will keep to largely miss the second one.
Laptops have outsold desktops every year for the last decade+. The M1, being a mobile friendly chip, is addressing the larger market. People may not care about power usage from the wall, but they do care about battery life.
Also, the fact that the M1 is even being mentioned along with a desktop powerhouse like the threadripper says it all. The M1 is the lowest end mobile Apple Silicon. We know higher end mobile chips are coming along with desktop ones. The M1 has put all the chip manufacturers on notice, much the same the way the iPhone did (to the point RIM thought Apple was lying about the iPhone because they thought the device was impossible).
Now if we look at what people are buying and ask ourselves 'are these the people who care about TDP?' the answer is simply 'no'
The majority of laptops sold are sub 700, as the previous post said, low budget, low power and in education, markets that Apple has never even considered (except education, but it's mainly college students in USA)
Not saying M1 aren't good, but that they won't shake the market as NVidia buying ARM could
P.s. M1 is a Zen 3 competitor, threadripper has been mentioned for the exact opposite reason: because that's the state of the art for people looking at performances in personal computing (while in HPC ARM is prominent)
It's only being mentioned alongside threadripper because the article went overboard and claimed the M1 "embarrass all other PCs — ... including ... every single machine running Windows or Linux"
This is a joke. The M1s performance is no where near the threadripper. Even on per-thread performance, new AMD cpus perform better than the M1... and they have many more cores. The M1 can only win these comparisons when we start adding all kinds of conditions, like power draw, heat, price point, etc. They need to add those conditions to redirect the conversation, because on raw performance threadripper destroys the M1.
The article made the mistake of leaving out those conditions to make it seem even more impressive than it is... but hyperbole and fantasy will always sound better than reality. Reality requires those conditions. That doesn't detract from the m1's impressiveness.. but it's foolish to forget that and go off and start making ridiculous claims like the article has done.
That was sort of my takeaway as well. I use desktop computers for most of my more demanding work. Laptops are great and (like smartphones, tablets, etc) any improvement in their capabilities is to be welcomed--but there's a difference between a headline that reads "New laptop is the fastest computer ever!" and one that reads "New laptop faster than all previous laptops!"
I don't do the most intensive tasks on thin, battery-powered computers for a reason. I try to do those on the box that's always plugged into mains power with heatsinks and fans as needed, plenty of ports to plug things in, and a big monitor. A more performant portable computer is awesome, but until I can spend a similar amount to make my workstation smaller without losing capability, I'll likely keep using it for heavy lifting alongside a cheaper, less powerful laptop or tablet for lighter work or remoting in.
You can't do what you do on a desktop on a laptop, not even a good one
Who cares if an M1 consumes less energy than a candle if I can buy 64GB of DDR4 3600 for 250 bucks and render the VFX for a 2 hours movie in 4k?
Another 300 bucks buy me a second GPU
When I deliver the job I put aside another 300 bucks and buy a third GPU
Or a better CPU
vertical products are an absolute waste of money when you chase the last bit of performance to save time (for you and your clients) and don't have the budget of Elon Musk
The M1 changes nothing in that space
Which is also a very lucrative space where every hour saved is an hour billed doing a new job instead of waiting to finish the last one to get paid
You can't mount your old gear on a rack and use it as a rendering node, plus you're paying for things you don't need: design, thermal constraints, a very expensive panel (a very good one, but still attached to the laptop body, and small)
So no, M1 is not comparable to a Threadripper, it's not even close, even if it consumes a lot more energy
When I'll see the same performances and freedom to upgrade in 20W chips, I will be the first one to buy them!
Then there's the 92% (actually 92.4%) of the remaining market that is not using an Apple computer that will keep buying non Apple hardware
Even if Apple doubled their market share, it would still be 15% Vs 85%
How is it possible that on HN people don't realise that 90 is much bigger than 10 and it's not a new laptop that will overturn the situation in a month is beyond me
And does it really matter to have a faster car if you can't use it to go camping with your family because space is limited?
That's what an Apple gives you, but it's not even a Ferrari, it's more like an Alfa Duetto
It's not expensive if you compare it to similar offers in the same category with the same constraints (which are artificially imposed on Macs like there's no other way to use a computer...)
But if you compare it to the vast amount of better configurations that the same money can buy, it is not
>You can't do what you do on a desktop on a laptop, not even a good one
Yeah… no, those days are over. The reviews clearly show the M1 Macs, including the MacBook Pro outperform most "desktops" at graphics-intensive tasks.
>So no, M1 is not comparable to a Threadripper, it's not even close, even if it consumes a lot more energy
Um… nobody is comparing an M1 Mac to a processor that often costs more than either the M1 Mac mini or MacBook Pro. However, the general consensus is the M1 outperforms PCs with mid to high-end GPUs and CPUs from Intel and AMD. Threadripper is a high-end, purpose build chip that can cost more than complete systems from most other companies, including Apple. However, it's at a cost of power consumption, special cooling in some cases, etc.
>Who cares if an M1 consumes less energy than a candle if I can buy 64GB of DDR4 3600 for 250 bucks and render the VFX for a 2 hours movie in 4k. Another 300 bucks buy me a second GPU
The MacBook Pro has faster LPDDR4X-4266 RAM on a 128-bit wide memory bus. The memory bandwidth maxes out at over 60 GB/s. And because the RAM, CPU and GPU (and all of the other units in the SoC) are all in the same die, memory is extremely fast.
From AnandTech; emphasis mine [1]:
"A single Firestorm achieves memory reads up to around 58GB/s, with memory writes coming in at 33-36GB/s. Most importantly, memory copies land in at 60 to 62GB/s depending if you’re using scalar or vector instructions. The fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before."
It can easily render a 2-hour 4k video unplugged in the background while you're doing other stuff. And when you're done, you’ll still have enough battery to last you until the next day if necessary. According to the AnandTech review [1], it blows away all other integrated GPUs and is even faster than several dedicated GPUs. That's not nothing; and these machines do it for less money.
>vertical products are an absolute waste of money when you chase the last bit of performance to save time (for you and your clients) and don't have the budget of Elon Musk
>The M1 changes nothing in that space
This is not correct… seeing should be believing.
Here's a video of 4k, 6k and 8k RED RAW files being rendered on an M1 Mac with 8 Gb of RAM, using DaVinci Resolve 17 [2]. Spoiler: while the 8k RAW file stuttered a little, once the preview resolution was reduced to only 4k, the playback was smooooth.
The M1 beats low end desktop GPUs from a couple of generations ago (~25% faster than the 1050ti and RX560 according to this benchmark [0]). Current high end GPUs are much faster than that (e.g the 3080 is ~5 times as powerful as a 1050ti).
Don't get me wrong - this is still very impressive with a ~20w combined! power draw under full load, but it definitely doesn't beat mid - high desktop GPUs.
(This is largely irrelevant for video encoding/decoding though as you can see - as that's mostly done either on the CPU or dedicated silicon living in either the CPU or the GPU that's separate from the main graphics processing cores.)
You're missing the point. I'm not trying to argue about which system is better, I'm just saying that the comment I'm replying to is saying incorrect things about GPU performance. I'll answer your question anyway though:
You could build a complete desktop system including a GPU that's more powerful than the one in the M1 for ~$1000, but certainly not a 3080. They're very expensive, and nobody has any in stock anyway.
An RX 580 or 1660 would probably be the right GPU with that budget. (Although you could go with something more powerful and skimp out on CPU and ram if you only cared about gaming performance).
- a 3080 costs > $750 . Good luck buying one, I would if it wasn't out of stock. On the other hand the gtx 1050 mobile that is on the M1 can be easily found on eBay for < $50
- yes, you totally can. The best thing is that with a 1k.entry level you can start working on real-life projects that have deadlines and start earning money that will let you upgrade your gear to the level you actually need, without having to buy an entire new machine. The old components can serve as spare parts or to build a second node. You don't waste a single penny on things you don't need.
Even though, it's true, you can't brag with friends that it absorbs only 20 watts full load and the heat of the aluminium body is actually pleasant
Besides the additional cores on the part of the CPUs and GPU, one main performance factor of the M1 that differs from the A14 is the fact that’s it’s running on a 128-bit memory bus rather than the mobile 64-bit bus. Across 8x 16-bit memory channels and at LPDDR4X-4266-class memory, this means the M1 hits a peak of 68.25GB/s memory bandwidth.
Later in the article:
Most importantly, memory copies land in at 60 to 62GB/s depending if you’re using scalar or vector instructions. The fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before.
Anandtech is comparing M1 vs A14. It's high performance for a cellphone part.
Dual channel DDR3L or DDR4L also has a 128 bit bus. 4200MHz DDR4 is clocked on the high side for most laptops, sure, but it's hardly unusual.
Run the numbers and you get the exact same throughput figure as for M1, which isn't surprising, because we're just taking width * rate = throughput.
So I'll repeat my assertion, downvotes be damned: the memory on the M1 is not special. The packaging and interconnect is interesting. It might reduce latency a little; it probably reduces power consumption a lot. But there's nothing special about it. The computer you're on right now probably has the same memory subsystem with different packaging.
Anandtech is comparing M1 vs A14. It's high performance for a cellphone part.
That’s where they started, but their conclusion was beyond that.
Did you miss the part where they said the fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before?
This isn’t only about A14 vs M1.
It’s not that LPDDR4X-4266-class memory is special; it’s been around for a while. What is special is that the RAM is part of SoC and due to the unified memory model, the CPU, GPU, Neural Engine and the other units have very fast access to the memory.
This is common for tablets and smartphones; it’s not common for general purpose laptops and desktops. And while Intel and AMD have added more functionality to their processors, they don’t have everything that’s part of the M1 system on a chip:
* Image Processing Unit (ISP)
* Digital Signal Processor (DSP)
* 16 core Neural Processing Unit (NPU)
* Video encoder/decoder
* Secure Enclave
There’s no other desktop such as the M1 Mac mini that combines all of these features with this level of performance at the price point of $699.
I don't think that's notable, sorry. I would expect that of any modern CPU.
> it’s not common for general purpose laptops and desktops
Well, yeah, because "memory on package" has major disadvantages. You (laptop/desktop manufacturer) are making minor gains in performance and power and need to buy a CPU which doesn't exist. Apple can do it, but they were already doing it for iPhone, and they must do it for iPhone to meet space constraints.
I think unified memory is the right way to go, long term, and that's a meaningful improvement. But as you point out, there is plenty of prior work there.
> they don’t have everything that’s part of the M1 system on a chip
They actually do! The 'CPU' part of an Intel CPU is vanishingly small these days. Most area is taken up with cache, GPU and hardware accelerators, such as... hardware video encode and decode, image processing, security and NN acceleration.
Most high-end Android cellphone SoCs have the same blocks. NVIDIA's SoCs have been shipping the same hardware blocks, with the same unified memory architecture, for at least four years. They all boot Ubuntu and give a desktop-like experience on a modern ARM ISA.
> There’s no other desktop ... at the price point of $699
I'm not saying any of those things. I'm saying a $1000 laptop from Apple is a pretty good value for a lot of people and I'm not sure you can do a lot better at that price point. When you include things like having an Apple store across town where you can go when things break, it pulls even further ahead.
The best that money can buy changes when you put the actual budget. Today, Apple may be the best that $1000 can buy.
> People buying Apple for its M1 low power consumption are a niche inside a niche.
Got it, watching Netflix in Chrome without the laptop getting hot is "a niche inside a niche" while GPU intensive 3d graphics workflows are mainstream.
If not getting hot while watching Netflix is the benchmark, I can show you a 400 euros laptop that can do 9 hours of Netflix on battery (or 7 hours of 1080p 60fps streaming)
Don't need no M1
AMD and Nvidia have sold more GPUs than Apple laptops by a large margin
This is doable on practically anything. Talking about being powerful with low heat/power consumption implies heavier workloads than Netflix because otherwise the "powerful" part is irrelevant.
Of course I'm sure you can find laptops with terrible cooling design that would get uncomfortably hot watching Netflix, but there's plenty of thin and light laptops available for < $1000 that would do just fine too.
> Not sure what shitty machine hes talking about, but my threadripper's are not impressed with some 4 core M1.
Gruber greatly exaggerates here. Particularly mentioning the Mac Pro in the same paragraph.
Apple's M1 isn't competing with higher end Intel/ AMD desktop class CPUs with external GPUs. The M1 is a laptop CPU, designed for good performance with good thermals and excellent battery life. It will be interesting to see what they come up with over the next couple years to replace the CPU in the iMac and Mac Pro. But that is next year.
While I dislike apple and have never owned any of their devices I can't help but feel thankful to them because now that we know this is possible soon others will re-create their success and we will hopefully have the same perks on a more open platform
I don't think you really pay that much extra for an Apple machine, especially for the M1 ones. I'm afraid that when/if fast ARM alternatives arrive, they will probably cost 80-100 % of what the MacBooks cost.
So basically, if you can't afford a MacBook today, you won't be able to afford a comparable high-end ARM offering from, say, Dell.
One big difference is that MacBooks (and Apple products in general) retain their value far better than any other computer brand. If you're looking at the used market to save money, a Dell looks way cheaper because it's depreciated far more than the Apple machine.
To be honest, I've never ever resold a laptop. Ever. I end up giving old laptops to friends or family who can't afford a new one. So resale price is a non-issue to me.
I honestly don't know about the M1's, but I've never liked the price/specs ratio of Intel macbooks. While it's true their build is solid and they look good, they always seemed overpriced to me regarding specs.
And even the build quality... a friend who's an Apple fan used to tell me "they are better quality, they have less problems" yet his macbooks seem to break down at about the same rate as my Dell, HP or Lenovo laptops, so...
I've decided macbooks look good, have good quality screens and the best touchpads. I'm unwilling to pay a premium for that, and I don't even like OS X/macOS...
> I'm afraid that when/if fast ARM alternatives arrive, they will probably cost 80-100 % of what the MacBooks cost.
Possibly! I wouldn't buy them when they are the latest new thing, either.
I feel this is not a very thoughtful comment. There can be many more reasons for not being able to afford a $1000 laptop other than local compensation right? Financial constraints like kids, mortgage, lack of spare cash for a new machine etc.?
First, this discussion is about the new MacBook Air that costs $1000 US which is why I mentioned that price point. I suppose tariffs and such would increase that substantially, so that's a fair point. My bad. That said, I believe there was discussion about the hypothetical situation where you did want a Mac.
More to the point, I checked your post history before I commented, and I gathered you are a developer, somewhere in South America (?). I suspect that the code I write is probably no better than the code you write. But your pay is presumably quite a bit less than mine. My point was that it's unfortunate that like work doesn't yield like rewards.
Now here I am going to make some assumptions... In all likelihood, people are amassing wealth off of your work in the same way they profit from mine. But due to local wage norms, they can pay you less and reap higher margins. They get to upgrade the yacht, but you might not even have enough free time time to use a boat even if you could afford one.
It's possible that I extrapolated too much and that your situation is nothing like what I have inferred. Sorry in advance if that is the case.
This is it. I'm excited because Apple just showed what is possible. Other companies are going to have to try to match it.
I have wanted a powerful "enough" fanless laptop for many years. I got excited about 4 years ago with Intel's short-lived M line and I bought an Asus ux305. It served me well for front end web dev (had to use a cooler stand under sustained high load though). Great Linux support and battery life (after installing tlp). I remember feeling a bit of awe that it was reasonably snappy with no fan, but it was not quite there. I can only imagine the feeling is an order of magnitude greater with the M1.
I don't own any Apple products, but I still hope that raising the bar will benefit me later on. I'll be pretty pumped when this level of hardware can run a mainline kernel.
>Other companies are going to have to try to match it.
This actually worries me.
Lately, anytime I look outside the Apple ecosystem, I wonder to myself what all these companies are doing... sure, I could see AMD closing the gap, but Apple's made additional choices for integration beyond what AMD would do alone.
I guess my summary would be: other companies trying to "match" Apple hasn't really gone well in the past decade, short of a few select standouts.
I've had same desire for a lightweight, fanless laptop for many years, also had (and still have and use) an Asus ux305, which I used for Linux development.
My main use machine now and for last couple years has been a Google Pixelbook. Lightweight (2.4 lbs) and fanless, solid 8+ hours on a charge, great for general use on chromeos, okay for Linux with built-in Crostini container.
I've had an iPad, but have never had any other Apple product. I've been pretty much anti-Apple since 1979, when as a 15-year-old I bought a Radio Shack TRS-80 over the much more expensive and glitzier Apple IIe.
I now have a Macbook Air M1 on order. We'll see how it goes. I'm sensitive to weight, prefer very light laptops. At 2.9 lbs it will be a noticeable step back from the ux305 (2.6 lbs) and Pixelbook (2.4 lbs). If I end up not bonding with the Macbook I plan to sell it; Apple's generally high resale value is a factor in giving the M1 a try.
But that won't happen unless there's a Windows ARM build that does everything that Windows does in x86: the market for Linux-only laptops is pretty small.
Work has already begun to port Linux to M1 Macs. Linux for ARM and Windows for ARM are already running as PoC on the M1 Macs with QEMU (running ARM-on-ARM).
I expect (very much hope) it will. I haven't contributed to the Linux on M1 fund yet, but I am seriously considering it, particularly if I get an M1 Mac. I likely wouldn't run Linux until after Apple retires MacOS for the M1, but want the option.
> What makes you think Windows won't take off on ARM?
I have no idea if it will or not. Microsoft's current strategy on ARM seems pretty unlikely to succeed though.
> In any case, if it can be done -- i.e. if there aren't locked down specs -- you can bet Linux will be ported to M1 Macs.
I'm confident (and hopeful) we will see an M series Linux port. How good it will be without support from Apple is questionable. The current state of Linux on Mac doesn't exactly inspire hope.
> Microsoft's current strategy on ARM seems pretty unlikely to succeed though
I'm not familiar with Microsoft's current ARM strategy, but they can always pivot of it turns out it's really the future. I must reluctantly admit Microsoft has been capable of some pretty surprising and welcome changes in its history.
Yeah, that would be really amazing. Is it even possible to install Linux on this new Mac though? Probably not. Will Apple offer these chips to other manufacturers? Would be cool if we could build small form factor PCs with these M1 chips!
I also wonder about the driver situation. Intel hardware has excellent support on Linux, everything works out of the box. Apple probably has zero interest in contributing drivers to the Linux kernel.
It's not just "ARM is fast," though—Apple's ARM chips in particular beat the pants off their ARM competitors. (See benchmarks of any iPhone compared to flagship Android devices.) I don't think you should expect similar levels of performance from a run-of-the-mill ARM chip on the desktop (though I'm happy to be shown benchmarks to the contrary).
It depends on what you're comparing: A14 on phones have a Multi Core score which would be quite comparable to Snapdragon 888 (extrapolating a 20% boost over SD865) that was just announced all while A14 being clocked higher (3.0ghz 4 core vs 2.84 x 1, 3 x 2.4) comparing performance cores. Single thread OTOH Apple wins comfortably but I would suspect X1 to reduce the lead a bit. As for AI performance Apple claimed 11 TOPs, while Qualcomm claims 25 TOPs
Apple's ARM SoCs are definitely ahead, but it's not like the competition is hugely far behind. Top of the line Android phones and tablets perform well enough for most desktop tasks IMO. It's mostly the software and I/O that makes them unusable as such.
x86 chips aren’t “unusable” either, and if these other ARM chips wanted to compete in the desktop space they’d be just as - or more - “hot and/or slow, clunky, noisy”. ISA just doesn’t matter half as much as microarchitecture.
To be clear I'm not talking benchmarks, I have no idea what the raw numbers are. I'm talking about my personal experience using ARM SoCs. Specifically I've worked with Tegra chip and they're powerhouses for instance, I could definitely see myself do my day-to-day development on such a chip.
When iPhone came out in 2007, Android was still an OS for BlackBerry like devices. It didn't take long for Android to discover touch screens once Apple made it clear that they were dramatically better. Similarly, it's not like other ARM chip manufacturers are going to have to go through all of the same R&D expense to figure out how to make chips that are comparably fast--they just need to copy what Apple did without infringing on any particular patents. Even if they did have to spend the same R&D expense to get there, Apple has already derisked it. So they have a small fraction of Apple's expense and none of the risk, so I would fully expect to see M1 competitive chips come out in a few years' time. This isn't to say that they'll catch up with Apple completely; only that they'll be able to recoup a significant amount of the gap in a relatively short period of time such that there will be a similar breakthrough in the PC market (this is effectively what TFA is predicting as well, if that lends me any credibility).
In my opinion, Apple "derisked" high-performance ARM chips with the A9-ish era five years ago. Other companies have had a long time to spend catching up to Apple mobile chips, and they haven't.
That's a completely reasonable challenge, and I might be wrong. My very amateur speculation is that, in the mobile space, Apple got incrementally better over time. Rank-and-file Android users didn't really notice the performance gap and thus didn't demand faster chips. M1 changed expectations for the laptop market virtually overnight, and rank-and-file users will expect M1 performance, battery life, etc. I think this represents an enormous market opportunity for PC manufacturers and thus chip suppliers that didn't really exist circa A9.
My key assumption is that there's some pareto principle at play--80% of the M1's performance gains over the next best generic ARM chip are due to a handful of key innovations while the remaining 20% come from a long tail of incremental improvements. This would mean that generic manufacturers could copy those handful of key innovations (modulo IP rights, which might be another fatal flaw in my speculation) and recoup 80% of that performance gap in just a few years. If that assumption is bad and the performance is entirely an accumulation of lots of minor innovations over time, then I suspect it will take chip manufacturers quite a lot longer to catch up.
There's more to it than chip design though. There is the OS work to take advantage of the chip's new features. And designing new chip features to accommodate sw bottlenecks. Apple has very close collaboration between software and hardware engineering, while in the competitive space, those two things are not under the same roof. Makes it tougher.
True, and I don’t expect generic manufacturers to completely close the gap between them and Apple, but I do expect there will be enough to copy that we’ll still see a step change in PC performance.
Qualcomm has been unable to compete with Apple's ARM phone processors for years. It's gotten to the point that the only processor that comes close to a new Apple SOC on benchmarks is Apple's previous SOC. Why do you expect the PC market will be different than the phone market?
> it's not like other ARM chip manufacturers are going to have to go through all of the same R&D expense to figure out how to make chips that are comparably fast--they just need to copy what Apple did without infringing on any particular patents
Very handwaivy but the devil is in the details of that "all they have to do". Copying success is not that straight forward or else AMD would have had a response to Conroe in 2007, Intel for Zen in 2018, and Qualcomm for A-chips around 2014.
> An early prototype had a close resemblance to a BlackBerry phone, with no touchscreen and a physical QWERTY keyboard, but the arrival of 2007's Apple iPhone meant that Android "had to go back to the drawing board".
I'm not sure how. iPhone was available in June 2007. Android went into public beta in November 2007, and reached 1.0 and was first released in a commercial device in September 2008. [1]
I guess Android, or what morphed into Android, was probably being developed, in a pre-beta state, when iPhone was announced in January 2007. Okay. It sounded to me like someone was saying that Android was released on Blackberry-like phones at the time iPhone came out. That was only thing i was objecting to. Sorry, a small point I almost didn't even bother to make, and probably not helpful to anybody, anyway.
Do they? Because i have yet to see any numbers that aren't clearly the result of a massively streamlined SoC. Yes the ARM chip is great, but the RAM is blazing, tiny, and more close to the ARM chip than anything out there for x86, right?
I'll be interested to see if Apple can keep this up with 128GiB of RAM and a beefy GPU.
So far to me M1 is great, ARM is great, but comparing a tiny amount of RAM soldered onto the CPU compared to flexible and huge RAM potential devices seems dishonest. Unless Apple manages to squeeze 128GiB of RAM on their chip, at least.
I want to see what my Blender workstation looks like on Apple's Architecture. Where i purpose built a huge CPU, GPU, and RAM for Blender rendering and large scenes. I suspect Apple will start seeing huge falloffs in the SoC design under larger workloads.
Apple is dominant in making phones though. And this Laptop is designed like a Phone. So yea, it's amazing for that.
Architectural analysis of the chip [0] suggests it is a massive improvement as a general purpose CPU, not just because of Apple's vertical integration and streamlining.
The magnitude of extra ROB entries, having an 8x wide instruction decoder, that massive L1I cache, and such fat cores with lots of ALU and other op units while simultaneously keeping over a 3GHz clock is a phenomenal architectural improvement overall.
Apple may not need to add more RAM locally. Other silicon designers like IBM with POWER10 are moving to serial memory with massive "L4" caches, seen previously with their Centaur memory controllers. I can see Apple using OMI or a similar serial memory bus to add extra memory, using 16/32/etc GB of PoP RAM for caching reads and writes to slower off-chip memory. I expect Intel and AMD will eventually move to a similar memory model, at least in server grade hardware.
Yea and to be clear i'm not saying it's not awesome. I'm just saying that a lot of things - like real world benchmarks of programs/etc - massively benefit from the proximity and speed of the RAM. Proximity seems to be a very difficult beast to scale. Works great in small scale, horrible in large scale.
Sidenote, even if Apple rolled out 128GiB tomorrow i probably couldn't afford it. As the OP says:
Hypothetically, with these new memory architectures, you could get the best of both worlds. Best case, you get a huge speed up if everything fits in the high-speed RAM, worst case, you fall back to the "old" RAM.
Side note, while Apple charges ridiculous prices for their upgrades, Mac Pros are actually some of the more ""reasonably priced"" computers comparatively. I've looked at HP and Lenovo's equiv. workstations, and you pretty much get screwed by everyone at that scale.
My desktop can take up to dual 128 GB sticks but one stick costs over $1,000, so I'm stuck with my 32 GB stick :(
... OTOH, it's pretty insane you can get 128GB of RAM for $1,000 nowadays. I remember when 8GB was a lot and cost several hundred. Time and miniaturization go on I guess...
It's just very close to the CPU, on the same package. It's not on the same die or anything. There are laptops that already have a shorter physical path to RAM than the M1 MacBooks, stacking the RAM right next to the CPU is common.
So it's just a matter of time before >16GB is available with an M1. These are just entry level machines in the first round.
I know I'm probably older than most people here, but the notion that 16 GiB is "a tiny amount of RAM" is tremendously amusing to someone who paid $740 in 1986 (with a developer's discount) for 2 "MB" of RAM (we didn't have MiB back then)...
Probably because there's basically no competition in the ARM space, especially in non-mobile. I'd like to see what amd and intel can do with ARM before saying apple will continue to be king.
They don't count in the PC space. For desktop PCs there's just the Mac mini and various hacker boards. For laptops there's the new MacBooks and the mediocre Qualcomm 8cx/SQ2 and nothing else.
For other ARM manufacturers, I wonder how much of this is a software problem.
I mean, what would the market be for a super-fast ARM laptop/desktop machine? It would run Windows RT, I guess?
How much better than Intel/AMD would the ARM machine need to be to justify buying it and running that? It seems like the M1's "magic" is not just the chip, but very much also the Rosetta thing.
Windows RT doesn't exist anymore. Now its Windows ARM, which is basically 100% fully featured Windows with the exception that x86 apps have to be run in emulation and x86-64 apps don't run at all (yet).
> For other ARM manufacturers, I wonder how much of this is a software problem.
IMO, one of the key innovations Apple made in the M1 is a modification to the memory model so that it is much more efficient to translate x86 code into ARM code.
There's lots of credible ARM work in both the mobile and the server space. Generally, a lot of it would work OK in the PC space, but for the lack of software support. Apple has the ability to leapfrog that particular problem.
Still, Apple seems to have done pretty well compared to both the other mobile and server ARM vendors with their PC offering.
This whole article and thread are about the desktop (not server) space.
I do agree that Apple's push to ARM for laptops will have a dramatic impact on the server space in a year or two. Intel is in for a world of hurt unless they can pull a rabbit out of their hat.
Dunno, seems more like now. Amazon has arm instances, there's a bunch of arm systems in the server space already and the world's fastest HPC cluster is already arm.
Samsung stopped Exynos development a year ago. Cavium is basically dead (yes, bought out by Marvell). Qualcomm is interested only in mobile SoCs at mobile SoCs volumes where they can also sell radios, and Nvidia was traditionally not capable of delivering anything at promised power consumption.
This is from wikipedia (under "Exynos"), so apply the appropriate amount of salt:
On October 1, 2019, rumors emerged that Samsung had laid off their custom CPU teams at SARC.[21][22][23] On November 1, 2019, Samsung filed a WARN letter with the Texas Workforce Commission, notifying of upcoming layoffs of their SARC CPU team and termination of their custom CPU development.[24] SARC and ACL will still continue development of custom SoC, AI, and GPU.[25]
You misunderstood what that means. Samsung is continuing to develop Exynos SoCs using stock Arm CPU cores and AMD GPU cores. This will require fewer people hence the layoff.
Exactly; they do not have their cores anymore, they are going to use ARM's Cortex. So their SoC will be always average at best, by definition. They won't be able to make anything exceptional.
HiSilicon's Kirin chips use Cortex microarchitecture and are very competitive. I didn't follow recent releases, but I remember that the Kirin 970 was way ahead of the Snapdragon and Exynos chips at the time, both in terms of performance and power efficiency.
Huawei often releases new Kirin SoC on October with Mate series meanwhile Qualcomm reseases flagship Snapdragon 8xx on December, so Kirin outperform latest Snapdragon that released on last year.
I think in the last year or so we hit an inflection point for ARM performance that makes it acceptable.
The Raspberry Pi 4 / Raspberry Pi 400 are halfway decent to use. The Lenovo Chromebook Duet is probably the first ARM-based Chromebook where I haven't felt constrained to keeping only three tabs open (and although it still struggles with some of the largest spreadsheets and dashboards at work, it is a delightful second-screen device).
But I think the M1 still has an edge because of the vertical integration between software and hardware -- if you previously wrote an app against Core Video or Metal, it should get hardware acceleration once you recompile it for the M1. Whereas if similar accelerators were exposed in Linux or Windows you'd probably have to write against a new custom API/library (such as nvenc).
I am not sure if the derogatory tone is fair. This is just example how progress works. Not that long time ago Apple had to be saved by some pennies thrown by Microsoft. Not that long time ago Apple was moving from Motorola processors to Intel ones.
There is nothing bad that someone stayed behind. Someone has to stay behind if we want progress. No reason to insult predecessors.
Planc and Dirac weren't making fun of Newton. Minitel was used in France as some simple version of Internet years before Internet was created. Was Minitel bad? Should we make fun of it?
Was Minitel standing in its underwear in front of Internet? Has Dirac stolen Newton pants?
Intel made a lot of great stuff. Now Apple is making great stuff (or, to be precise, TSMC does). Next year someone else is going to make great stuff. Maybe Apple maybe not.
Intel and AMD chips are not that far behind, depending on what benchmark is used, even though they use older 14/10 or 7nm technology.
I also really hate the concept of not being able to upgrade RAM - If I use Mac and I need more RAM I need to throw away my old Apple laptop and buy a new one. So I have to spend way more money, not to mention impact on environment.
So, today, I am still choosing Intel/AMD computer, I can install whatever operating system I wish and upgrade my system to whatever configuration I dream about and I can afford now or in future.
I agree with you but I basically filtered that as "typical Mac fanboy noise" and try to search for insightful content which I think this piece has, in spite of its "I have a shrine dedicated to Steve Jobs in my living room" tone.
The point about being able to upgrade ram or upgrade to whatever you want doesn’t make sense.
A ram upgrade will never compensate for a less efficient memory architecture.
There is no CPU architecture other than Apple’s that will make your laptop faster because AMD and Intel are thermally limited, so there is nothing to upgrade to.
The only situation in which you are right, is if you are talking about a high end desktop tower.
>The point about being able to upgrade ram or upgrade to whatever you want doesn’t make sense.
If my program needs 64GB of ram, your 8GB highly efficient "memory architecture" can do nothing about it...you have to know that even my gpu has 16GB alone. But i have to say that 8GB are probably enough for your Chrome Browser.
It's actually really impressive how well Intel desktop CPUs still do considering they're still made on a (continuously improved) five year old process.
> Not that long time ago Apple had to be saved by some pennies thrown by Microsoft.
Agree largely with the rest of your comment, but this particular event was more than 23 years ago. I'm sure a good number of readers of this thread is younger than that!
> I think it's a bit unfair to lump Linux in there since it's been running on ARM and other embedded, low-power devices basically forever, but it is true that few of us run our Linux desktop on an ARM board.
We aren't running ARM on our desktops because there aren't ARM desktops for us to use. The market jumps from low-end boards like the Pi and Pine's lineups - which are excellent for what they are, but suffer from poor performance and I/O - to the high end where ex. https://www.asacomputers.com/Cavium-ThunderX-ARM.html looks like a nice system but $2,820 is a bit of a barrier to entry.
Exactly! I would love to switch my systems off of x86, but I can't buy anything performant for a price that I'm willing to spend on hardware. I might yet try a Pi 4 as a desktop with USB for disk, but that's hacky and still suffers from performance problems. Or, I could get something with good performance, either POWER or ARM, but then I'd have to pay through the nose for the privilege. Now, some of this is market size: I get my x86 boxes at a huge discount by buying used, and high-end ARM/POWER machines aren't popular enough to really show up on eBay (yet). But in practice, there's just no good options.
The thing is not just "ARM", it's Apple's implementation. Their A* mobile cpu lineup destroys any competing mobile ARM CPU, and the M1 evolved from that. So just running ARM isn't going to save anyone.
> But I think the overarching point is true: now desktop and laptop makers can't just pretend that having a hot and/or slow, clunky, noisy x86-based architecture is just a fact of life. I hope and expect that it'll help create a new generation of ARM-based laptops and maybe even desktops that will run cool and smooth.
Be careful what you wish for. X86 is the most open architecture right now. ARM, on the other hand, has locked bootloaders, messy device trees and so on. You can see that virtually every Linux distro running on an ARM SOC needs some specific tweaks.
I know very little about this space, but presumably that will change if/when a significant portion of the laptop space (and probably the server space as well) moves from x86 to ARM? It might be painful for early adopters for a few years, but I would expect this stuff to get ironed out in time, no?
No, that ship has sailed. Anything new that comes now can be built according to the expectations smartphones have set over the past decade and a half. Closed systems that are treated like disposable items. Even PCs in their mobile incarnation are now far less open and flexible than they used to be.
The reason PCs are still open is historical, they started like this and people can't really see PCs any other way but they're a dying breed and whatever comes after is a different game with different rules. We'll have some niche manufacturer building something more open for the small segment of enthusiasts.
There's a difference between swapping or adding any of the many standardized components you want or installing pretty much any operating system you can think of on your PC, compared to enabling "developer mode" on an Oculus Quest.
The fact that so many people casually say how a platform is open or flexible because if you jump through enough hoops you get to make some minor modifications to your device, or applauding the openness of a system for allowing you to sideload standard apps is the reason manufacturers don't see a need to propagate the PC philosophy anymore. Too many users grew up with this new model so they don't see the loss.
None of the devices you mention are user serviceable in any way hardware-wise, the software is almost always locked onto the device and there's no easy, or even supported way of changing it. And if you do change it you just get a small variation of the same software. Your phone may get a slightly degooglified version of Android but it's unlikely you'll get it to run Ubuntu, Windows, or a FreeBSD without a massive investment in time and knowledge.
My understanding is that developer mode on Android and Chromebook will unlock the bootloader. Software wise, that's about as user-hackable as you can get. It's true that the Oculus Quest is not as open.
I don't expect or even really want to swap hardware on these devices: if that was a priority for me I would get a PC, Arduino, Raspberry Pi, etc. I get the desire to tinker and customize, but from where I'm sitting I have plenty of opportunity to do that.
I also don't expect the device manufacturers to do the work to make Ubuntu or FreeBSD run. They didn't do that for PCs, the community did.
> Be careful what you wish for. X86 is the most open architecture right now
You might want to take a look at PowerPC/OpenPOWER.
ARM not being very open is mostly because there's been limited development interest in it, aside from specialised devices, and those companies have no real need or interest to have open things. Doesn't mean ARM itself can't evolve past it.
x86 is not "Open" in any fashion. You can't license x86 even for a fee, let alone free.
While ARM isn't open either, it is by comparison far more available to license at reasonable cost.
> ARM, on the other hand, has locked bootloaders
It also has open/ unlocked boot loaders. One of the advantages of licensable technology like this is you can have a lot of different architectures with the same underlying core. With x86, you get whatever the Intel/ AMD duopoly decide to ship you.
> x86 is not "Open" in any fashion. You can't license x86 even for a fee, let alone free.
OP is not talking about being able to manufacture x86 CPUs kind of open, that's useless for most folks.
They're talking about being able to install any OS, drivers being fairly universal, no device specific out of tree kernel patches for the vast majority of PCs, no device trees etc.
It's a lousy use of the word "Open" in a discipline where Open has a fairly specific meaning.
I mentioned that as well though. Using your goofy definition of Open, x86 is no more "Open" than ARM. Pinebook and Raspberry Pi both use no bullshit ARM setups. With ARM, how accessible the architecture is is 100% dependent on the implementation which vary greatly by manufacturer.
If I buy any x86 PC, chances are I can install mainline Linux on it. Picking up any particular recent ARM SoC and the chance of doing so is much, much less. The M1 included, after all.
> Pinebook and Raspberry Pi both use no bullshit ARM setups.
Both of these took considerably longer to get mainline support that your average x86 hardware needs.
> Both of these took considerably longer to get mainline support that your average x86 hardware needs.
Not sure what you are thinking here. It's taken 25 years so far and we still have tons of x86 hardware which isn't supported well in Linux. There is a reason Dell ships a specific laptop for developers with Linux pre-installed.
Consider for a moment how many independent developers rely on x86 based Linux systems versus ARM. x86 isn't open, Linux's support for it is a symptom of Intel's monopoly.
> And yet, even Windows DELL laptops are more likely to run mainline Linux than any average ARM SoC, even including the ones targeted at developers.
This is more about inertia and developer resources than openness. The more developers you have on ARM, the faster device and software support will come.
> I am no fan of x86, but trading what we have now for an M1 is a net loss in terms of our ability to install anything we want as users, no question.
You've changed the question here. We were speaking about ARM in general, not the M1.
Well there's two issues there. One is if the bootloader is open, which it is.
The second is if there's linux port for that hardware, which it isn't. There's a project underway, seems likely that the CPU (which is still AARCH64) will work, and likely networking. However to make a usable desktop you'll need a GPU which is much harder.
> x86 is no more "Open" than ARM. Pinebook and Raspberry Pi both use no bullshit ARM setups.
They're ARM SoCs, meaning that you can't just download the Ubuntu ARM ISO[1] and run it on either device like you could with an x86 ISO and an x86 machine.
To run Ubuntu or any Linux distribution, you need to use images that were specially crafted for either SoC.
This is because ARM SoCs use custom bootloaders and don't have hardware on enumerable buses. Each SoC model is essentially a unique hardware configuration that needs an individual Linux port.
ARM servers are more similar to x86 machines in that they use UEFI and have enumerable buses.
As long as machines are shipped that use ARM SoCs, Linux support for all of them will be quite difficult to achieve[2] compared to x86 machines.
While Apple likely has no motivation currently to make a UEFI-based ARM computer, there's nothing that prevent anyone else from doing so. UEFI is not x86-specific.
(I also think it's a bit too early to say definitively that no ARM-based Apple computer will ever be able to boot into Linux. I don't think Apple has any motivation to make that easy, but that may not be the same as making it impossible.)
It's cheaper to just design a ARM SoC with a custom bootloader.
> I also think it's a bit too early to say definitively that no ARM-based Apple computer will ever be able to boot into Linux.
There's a big difference between a SoC running a Linux fork and actually running Linux distributions in the same sense that we can on x86 computers or ARM servers.
Windows Mobile/phones actually had UEFI bootloaders with Windows Boot Manager [1]. For some Windows mobile devices, people have managed to unlock the bootloaders and load other EFI applications.
UEFI can be just as closed as custom non-UEFI bootloaders. Windows Mobile devices were UEFI, but they had Secure Boot enabled making it almost impossible to boot non-OEM signed images. (Un)fortunately, their Golden Key was leaked allowing experimentation with some Windows Mobile devices [1].
Compared to the amount of bullshit we had to to deal with in the 90s and 00s getting Linux running on a laptop, it's a walk in the park. Often it was a choice between mediocre FLOSS drivers or binary blobs from 2-3 private parties... when things even had drivers at all. It's been a while, but how many laptops lack wifi support or require proprietary drivers to work now?
Are the FLOSS drivers for Nvidia's cards great now? How many years have people been hacking on that project to make them work as well as they do?
What people are describing as "Open" is actually an enormous amount of work on the part of developers and the grudging support of hardware vendors after decades of struggle.
> It's been a while, but how many laptops lack wifi support or require proprietary drivers to work now?
Very few actually.
> Are the FLOSS drivers for Nvidia's cards great now?
Don't shift goalposts, we are talking about CPU support, not GPUs... GPU support on ARM is anyway not that great to begin with.
> What people are describing as "Open" is actually an enormous amount of work on the part of developers and the grudging support of hardware vendors after decades of struggle.
Intel has Linux dedicated teams at least, pushing FOSS code every single day. Most ARM licensees could not care one bit about FOSS.
> Don't shift goalposts, we are talking about CPU support
Linux already supports ARM CPUs and has for some time. The challenge is drivers. Getting video/ wifi/ etc running is the part that makes devices actually useful. The Broadcom Blob the poster I replied to was referring to is not for supporting ARM. Very similar to the Nvidia blob many people install (often unknowingly) when they install Linux on a system with Nvidia.
> Intel has Linux dedicated teams at least, pushing FOSS code every single day. Most ARM licensees could not care one bit about FOSS.
Intel didn't give a shit about Linux for how many years? Companies don't care about FOSS until it has potential to affect their bottom line. It was only after Linux servers were running half the Internet, that Intel stepped up and started supporting Linux.
I'm guessing you missed the first couple decades of Linux adoption. Linux on ARM is just a fast motion repeat of the early days of Linux on x86.
They don't seem that expensive. They're high-end machines themselves, and their pricing seems on-par with high-end x86 workstations, like the Dell Precision series.
> * the x86 world resolves around standards such as ACPI, BIOSes and general-purpose dynamic buses.
> * ACPI normalises every single piece of hardware from the perspective of most low-level peripherals.
> * the BIOS also helps in that normalisation. DOS INT33 is the classic one i remember.
> * the general-purpose dynamic buses include:
- USB and its speed variants (self-describing peripherals)
- PCI and its derivatives (self-describing peripherals)
- SATA and its speed variants (self-describing peripherals)
> exceptions to the above include i2c (unusual, and taken care of by i2c-sensors, which uses good heuristics to "probe" devices from userspace) and the ISA bus and its derivatives such as Compact Flash and IDE. even PCMCIA got sufficient advances to auto-identify devices from userspace at runtime.
> so as a general rule, supporting a new x86-based piece of hardware is a piece of piss. get datasheet or reverse-engineer, drop it in, it's got BIOS, ACPI, USB, PCIe, SATA, wow big deal, job done. also as a general rule, hardware that conforms to x86-motherboard-like layouts such as the various powerpc architectures are along the same lines.
> so here, device tree is a real easy thing to add, and to some extent a "nice-to-have". i.e. it's not really essential to have device tree on top of something where 99% of the peripherals can describe themselves dynamically over their bus architecture when they're plugged in!
> now let's look at the ARM world.
> * is there a BIOS? no. so all the boot-up procedures including ultra-low-level stuff like DDR3 RAM timings initialisation, which is normally the job of the BIOS - must be taken care of BY YOU (usually in u-boot) and it must be done SPECIFICALLY CUSTOMISED EACH AND EVERY SINGLE TIME FOR EVERY SINGLE SPECIFIC HARDWARE COMBINATION.
> * is there ACPI present? no. so anything related to power
management, fans (if there are any), temperature detection (if there is any), all of that must be taken care of BY YOU.
> * what about the devices? here's where it becomes absolute hell on earth as far as attempting to "streamline" the linux kernel into a "one size fits all" monolithic package.
> the classic example i give here is the HTC Universal, which was a device that, after 3 years of dedicated reverse-engineering, finally had fully-working hardware with the exception of write to its on-board NAND. the reason for the complexity is in the hardware design, where not even 110 GPIO pins of the PXA270 were enough to cover all of the peripherals, so they had to use a custom ASIC with an additional 64 GPIO pins. it turned out that that wasn't enough either, so in desperation the designers used the 16 GPIO pins of the Ericsson 3G Radio ROM, in order to do basic things like switch on the camera flash LED.
> the point is: each device that's designed using an ARM processor is COMPLETELY AND UTTERLY DIFFERENT from any other device in the world.
ARM UEFI systems exist. There are now the norm on servers along with ACPI. Even some embedded devices use UEFI. There are other standards to address those various points.
This post is so outdated, that ARM servers' companies had the time to have everything, then still die (most of them). And then revive with the AWS Graviton2 and other Neoverse chips.
It's not outdated, it holds true for ARM SoCs. Consumer ARM devices are ARM SoCs, and not server hardware. It's cheaper and easier to design custom SoCs than it is to build consumer hardware that conforms to the standards that ARM servers do.
The apple M1's don't have locked bootloaders, there's a warning in the firmware about enabling unsigned booting, but it's just warning. That's why there's an attempted linux port underway.
The whole narrative has changed over the last few weeks too.
"Fastest chip" uhh well no..
"Fast for this one thing", yeah it's 0.1s faster opening that program no one ever said was slow.
"It has good battery life"
So this is the new narrative? Are nerds constantly fighting back against Apple marketing and they are adapting?
Usually their marketing team just insists over and over. Maybe because this is actually measurable (unlike privacy and security) their legal team is holding the marketing team back.
Such a silly argument. An M1 Mac embarrasses no one but prior Macbooks, as it only runs OS X. There is no useful comparison to Windows or Linux. It's as asinine as comparing a mobile phone performance to a desktop PC.
Because the software I use that works on Linux or Windows doesn't exist or cannot be run on OS X. CUDA doesn't even support OS X anymore. If I can't run my OS/software of choice, it doesn't matter how great M1 is.
I think my original response was a bit overly dramatic. But my point still remains. I don't think you can compare an OS X based laptop to anything else. The software available is far too different, and OS X is much more locked down (and getting even more locked down).
It is fair to lump Linux in there. Linux is supported on pretty much every largish Arm system other than Apples, at least if you include Android Linux. No other Arm system comes close to matching the M1 in terms of performance.
I don't think that is an indictment of Linux. It is likely if the specs for the M1 were released, Linux could wring more performance out of the M1 than MacOS. At least in some workloads.
I don't understand this _extremely_ common viewpoint: Intel has been putting out awesome fanless chips for Windows and ChromeOS for a couple years now, and Tiger Lake scores ~1400 single threaded/3700 multithread on 4 cores in GeekBench. M1 is __awesome__ but it's exactly what you'd expect from a Tiger Lake die shrink, which is coincidentally, the same die M1 is on.
I wish Gruber would just let modest criticisms like this stuff fly rather than pen an overwrought argument like this.
I think it’s perfectly legitimate to raise concerns around Apple software because it’s one area where Apple has faced sustained criticism from its users and has done very little to remedy the situation. So if the software that runs on top of the M1 is shakey, it’s entirely realistic to expect these “warts” to be with the platform for a while.
Eh, I get his coming down on criticisms this time. It doesn't make sense to criticize the hardware for software issues. I'm not going copy a binary compiled for x86 over to a RPi and complain about ARM chips being inadequate.
If there are software issues, address them as such.
I think a lot of his frustration seems to come from people using Apple software to write off ARM hardware overall. In the M1 specifically the software may hold it back, but using that to justify claims of x86 systems being fundamentally superior.
The justification that 16GB is "enough" from the Apple blogger clique like Gruber is ridiculous. Even if it is "enough" it's still what my MacBook Pro I bought 6 years ago shipped with for several hundred dollars less than this one.
Why are we giving Apple a pass on RAM being stuck at 8GB and SSD being stuck at 128GB (only very recently bumping to 256) for almost a decade. Not to mention some of us get stuck on 8GB 128GB machines because our workplaces only buy the base "Pro" machines, refusing the drop the extra few hundred to make it what Gruber considers the base level.
SSDs are no longer expensive, go on Amazon and see how much 1TB M2s cost these days.
Some people wish they could get the performance of M1 and 32GB RAM and four Thunderbolt ports. They will get it eventually but it's a shame it's not available yet.
I've always gone cheap for personal use and the last 3 MacBooks Pros I've purchased were i3/i5 8GB models, and never have had perf/memory issues (even with 2 users logged in)
Work provides i7/i9 with maxed out RAM and honestly don't see a night and day difference in performance.
I want a "hot" M1, one that runs at 5Ghz and burns 120W (or whatever) at load because it will be even faster. That is how this works, the reason your seeing "fast and cool" is because the target is 15W. If that becomes 25W then there is budget for more cores, or faster clocks (or whatever) and the the result is a faster machine. Which is why a lot of people are holding out for the M2 (again or whatever) the next version they put in the bigger machines because it should be even more impressive if apple can scale it.
I guess that is what is coming: M2 will be 40% faster, still be 10W but come in 1x for Air and 2x configuration for MBP 16 and desktop Macs. I think this is just the beginning of a growing wave lapping the x86 space.
I am willing to bet if mods did some data munging they would find sockpuppetry of a hail corporate type at work, because if hn is really this brainwashed about macs it makes me sad.
Apple has always operated with a more closed-off ecosystem that sacrifices openness for tighter control on the user experience, and that's always been at odds with the "I want access to everything so I can tinker" mindset on HN. I expect any HN discussion about Apple to be a mess, because the average HN commenter/reader wants something very different than what Apple sells.
I think something like 80% of Silicon Valley tech workers are using Macs and tons of them hang out on HN. I have always seen the HN crowd as being mostly pro-Apple.
Maybe I'm wrong though - it might depend on the thread. Whenever I put out an anti-Apple sentiment on a pro-Apple article thread I usually get massive amounts of disagreement.
My own dislike for Apple goes way beyond their protectionist policies. Maybe if I didn't hate the Mac UI with that hideous global menu bar and horrid window management, I could be more forgiving towards their politics. Heck, I used Windows as my main OS for decades until I finally switched to Manjaro and XFCE for the sheer convenience of running Docker and npm/yarn smoothly without all the Windows cruft slowing me down all the time.
It's funny though - I do less tinkering on my Linux desktop than I ever did on Windows (or my Mac), since everything just works the way I want it to in Manjaro. I think my first week using it, I had to figure out how to write an xdotools script to get window snapping in XFCE. Never had to touch it since. My Mac has broken things like Karabiner, Better Touch Tool, Little Snitch and so forth after many updates - so much so that I just stopped reinstall them. Luckily, I only have to use it once in a while for doing the occasional Mac/iOS thing.
I don't doubt Apple's CPUs are great, but they coincide with the Mac's transition into a product I, otherwise, no longer wish to use.
I want access to my boot volume. I don't want cloud anything.
So Apple's technical prowess now makes me sad. I can either use an OS that doesn't head further every year in a direction I hate, or I can have a fast computer.
Are M1 macs faster by significant factor for dev purposes though? I remember reading compiles are slow on MacOs, but don't remember what was the reason.
I'm also pretty bummed that I can't use the magic ARM machine yet, but this is just the beginning. Apple currently has an advantage, but I suspect its more of a market advantage than a technical advantage. Snapdragons are getting really fast too, and a lot of the difference in power comes down to a lack of demand. Now that ARM has been shown to be this good, more and similarly powered chips will be made.
> I actually discovered that I’ve had an instinct of measuring my MacBook’s CPU usage by feeling the heat on the strip of aluminum right above the Touch Bar, and I can’t even do that anymore now.
Market opportunity here for a peripheral that's just an adhesive strip you put above your touch bar, and it outputs heat corresponding to CPU usage ;)
The fact that reviewers of M1 chips are also saying things like "Safari is faster than Chrome" or "iOS is faster than Android" do make the other statements on M1 lose credibility.
"Safari is faster than Chrome" NOPE
Now that IE is gone, Safari is the new IE by being the last not-ever-green browser always behind in features.
So many benchmarks again and again proved Chrome and Firefox to be faster than Safari.
"iOS is faster than Android" NOPE
It just depends on the device. Of course the latest iPhone is faster than a 300$ mid-low range Android device.
Pick the top Android devices vs iPhones and you will see it's a tough race with no clear winner.
1. They took all the different chips and crammed more of them onto one chip. Now the different parts of the computer can talk to each other faster without any more work.
2. The new chip does lots of little things waaaay faster than the old chip does a handful of big things.
3. Apple figured out how to do a lot more things before they need to. Like if you did all of the math problems in your textbook ahead of time and then when the teacher eventually gives you the assignment you hand in the ones they asked for.
The processor looks at the future instructions it will run, can see that some things can be run ahead of time, and can run multiple multiple instructions in parallel (on the same core). You could run individual multiplications in parallel. Most of those things other ARM processors can do that too, but M1 is bigger in some of those sizes, like how many instructions ahead it looks at, how many things can run at once etc.
Which other ARM chips are you comparing with? It’s not much faster than A14, Apple’s other ARM chip. Which in turn is around 20-30% faster than A13, etc. Apple has been making their ARM chips faster very gradually over the last 8 years. M1 is just the first one they’ve put in a computer and more people are paying attention to their speed.
881 comments
[ 2.6 ms ] story [ 491 ms ] threadComing to the M1 laptops, they beat comparable Windows laptops in every single metric. It's literally just a matter of looking at the numbers and realizing how M1 has every Intel chip beat (in this category).
In fact this makes me wonder: if Apple chips are so far ahead, why are iphones not the fastest, best smartphones by a wide margin?
AFAIK, iPhones do have the fastest benchmarks (and the smoothest experience) across all phones.
I'm actually thinking about getting one of these as I've been without a personal laptop for a while now. That said, I'm waiting for Docker support to get flushed out a bit more as I wouldn't mind getting work done. I am interested on how the really high performance ram effects my workflows, as I've often gone over 16gb as a sticking point, got my work laptop upgraded to 32gb earlier in the year and my personal desktop is at 64gb.
I'm not sure it will replace my r9-3950X desktop for getting things done, but it's absolutely compelling for its' space. I swore off macs for a while now, and this might pull me back in.
Overall experience feels better on android. Few things that make difference for me.
1. Notifications are leagues ahead of anything on iOS.
2. Little accessibility features like sound search, automatic caption, text selection from any screen/pictures, better integration with google assistant adds up.
3. Customizability. Yes, even now iPhone is super limited in the layout you can have on home screen or the new app launcher. You can't group things or put them wherever you want on the screen. This is ridiculous.
4. Android phones have higher refresh rates. Almost any flagship in 2020. The punch hole camera feels better than the notch. Face ID = problematic in the pandemic with masks. Though, I like it generally but apple could have given touch ID on the same device as well.
5. For tinkerers, it has better support again. Youtube vanced, tachiyomi, advanced adblockers, "real" firefox with extensions, etc are only available on android.
The best part is you can get the pixel 4a for $300 with some carriers. 3 years of updates which is what typical upgrade cycle looks like for iPhone users even if they get "updates" for longer.
Source: Own both.
I think notifications are perfectly fine on iPhones. I've not used android for 4 years now, but if their notifications are the same as they were in 2016 (or just marginally different) then I don't miss much. I confess point 2 sounds very appealing. I do use custom DNS for adblocking, so I don't think I miss it on the iPhone, and tbh I've fallen in the rabbit hole of tinkering with my phone enough that I feel more productive without a phone I can tinker with.
Some apps are better on iOS but it's not a big difference because some apps are better on android as well. People care about different apps in the end.
There was a point in time that I was pretty deep into the Android scene, but at this point in my life I just want my phone to work consistently and have great battery life.
After many years of using Android devices (including many Nexus / Pixel devices), I switched to an iPhone 11 Pro Max last year. I've been very happy with that switch, my phone always makes it through a day of use and I haven't ran into any of the quirks that I encountered on my Android devices on a regular basis.
I completely understand why you would prefer Android, but there are definitely people (including myself) that are happier with iPhones.
I just don't think any flagship in 2020 is worth it for an average consumer when options like pixel 4a exists. iPhone se is not good due to the screen but otherwise it would have been a great contender as well. People notice 90hz crisp display more than faster opening apps.
Also I fundamentally will not buy a piece of hardware that will cease receiving security updates within 36 months of purchase. That's insane, and frankly a major fuck you to anyone considering buying your product.
Of course iPhone and Android can do all the same tasks. One just does it a lot better efficiency and power wise.
How does the chip performance matter if the Note feels faster in opening apps and scrolling due to the refresh rate?
How does efficiency matter if the Note holds more apps in memory for faster task switching?
How does a faster chip matter if the image processing is way superior in the Note so your photos almost always come out better?
The difference is subjective. I would have expected it to be objective like in the M1 laptop. The new air and pro absolutely obliterate other comparable systems in every metric.
For example the iPhones don't have the longest battery, but the phone that has the longest battery has two times the amount of battery, while having a much higher consumption than iPhones. If Apple decides to stick a 5000+ mah battery in an iPhone - Android phones would have no way of catching up to its battery life.
Another big thing being the latest screen tech - Apple can't afford to do an OLED screen on the iPad, since Samsung and LG can't possibly manufacture enough OLED screens to meet demand. Where Samsung is perfectly fine doing it, when they won't sell enough tablets to matter.
In the end it's not that the iPhone is clearly inferior to Androids, it's mostly that Apple has some numbers in mind that they consider optimal and going beyond those won't increase their sales (certain amount of battery, certain screens, etc).
It is also very common that people that like Android are not aware of this for some reason.
Sometimes this means the software is more clunky, less polished and isn't as fast... but I'm still using my now 3yo Pixel 2XL, which is still mostly acceptable (I really need a new battery).
Vivo Iqoo 5 Pro 671,218
OnePlus 8 Pro 590,112
iPhone 12 Pro 579,778
As ever, it depends what you're measuring
Apple has pretty consistently had the fastest SoCs for phones and tablets for a few years now, particularly in single core performance. Here are some benchmark results of the iPhone 12 versus some of it’s Android competitors [1]. Notice that last year’s iPhone 11 Pro Max still outperforms Android phones from this year as well.
I think part of the reason this isn’t a major differentiator (for phones at least) is that phones have been “fast enough” for several years now. Given the very aggressive throttling that is utilized on these types of devices, I think the primary difference ends up being battery life / efficiency rather than raw performance.
It’s also worth noting that while the performance / efficiency gap between x86 and ARM seems to be pretty sizable at this point, the gap between various current generation ARM CPUs isn’t nearly as drastic.
1: https://www.tomsguide.com/news/iphone-12-benchmarks-this-des...
I think for some people (eg my oldest brother), the iPhone is "the best." He's not a tech person - and that might be perspective that you're missing.
I tend to agree that there are great androids that seem competitive with iPhones (disclosure: I haven't really owned a flagship phone for a while; this is one of android's issues! The flagships aren't that much better than the $300-400 phones)
But I know a lot of people who think the iPhone is the premium phone experience, and if they're going to spend over a thousand dollars on a phone, it'll be an iPhone.
My point here is that for many people, the iPhone is objectively better.
The power/performance ratio of the M1 SoC does not seem to need religious adoration to be seen as a significant step in the establishment of ARM as a real competitor to the AMD64 hegemony. The wait is now for other vendors - Samsung, Qualcomm, etc - to launch similar "desktop-oriented" SoC's which can be used by both traditional "dekstop" vendors (HP, Dell, Lenovo and to a limited extent Samsung etc) as well as traditional mobile-oriented vendors like Xiaomi. Some of those vendors will eventually produce ARM-based platforms which support user-upgradeable memory, GPUs and storage - and possibly also user-upgradeable CPUs - like they do for AMD64-based systems. Once these systems become available ARM has a real chance to take over the market.
I'm very excited about the new m1. While I'm tempted to buy it, I also can't wait to see what they do with it in their pro models. Or what the second generation looks like.
The competition is great for everyone, and Apple going all in on arm as they run from Intel's manufacturing failures is very well-timed.
I think this was somewhat inevitable, to be sure. Microsoft had Qualcomm and AMD processors in their surface [2], and I think as TSMC continues to dominate in manufacturing, we will see really exciting gains.
Also, I'm thrilled that CPUs are coming with tensor cores now. I think fast matrix operations might be separated from gpus (at the least, you no longer need a >$400 gpu for it), and that's a win for everyone, too.
In two years, we could be able to reimplement early neural network research on a macbook air! (I mean, you already can, but instead of taking minutes/hours/days, even more recent things become accessible)
[1] https://www.engadget.com/2011-05-17-bbc-loving-apple-looks-l...
[2] https://www.theverge.com/2019/10/2/20888999/microsoft-surfac...
iPhones do likely get better longevity out of their faster hardware, mind you.
It's really not, though. You have to compare performance at a given power target to judge performance/efficiency, which you can't do for most ARM CPUs. But you can get 8 core x86 CPUs that are 15W, or 1.8w per core (example: the 4800U). Similarly 64 core Epycs are around or less than 3W per core. These are all well within similar per-core power numbers as your typical big-core ARM chip.
The perceived gap is larger than it really is, because there just isn't really an x86 tablet-focused market. The M1's jump into laptop-space puts it up and we can really see it shine, but the other ARM jumps into laptop-space were a joke. So comparing equivalent power targets becomes challenging.
But take your articles benchmarks for example. The Snapdragon manages something like 60% of the performance of the A12's. So take the existing M1 mac, cut the performance drastically, and suddenly it'd be not interesting at all. The performance/efficiency wouldn't be impressive at all if it was getting 60% of the performance it is now.
Because articles like that are pure nonsense.
Take the initial paragraphs. We already knew that faster processors don't necessarily run hotter. People are simply conflating higher clock speeds with higher performance, while ignoring all of the ways CPU design optimizes computation. Likewise, we already knew that "emulation" doesn't imply that performance will be unbearably slow. There has been enormous amounts of research and development into the problem over the decades, to the point where much of our software is optimized at runtime.
None of this is meant to diminish what Apple has accomplished. They were clearly paying attention to details that have been largely ignored or at the very least rarely highlighted by their competitors. Now that it has happened, we will likely see more from the PC industry. Apple will likely have a hard time keeping up unless they have more up their sleeve.
Apple’s Rosetta 2 doesn't do emulation; it translates the x86 code into ARM code on the fly and runs it, which is why most apps run faster on M1 Macs than they do on Intel Macs--and comparably priced PCs.
Now that it has happened, we will likely see more from the PC industry.
Probably not. If Intel and AMD were capable of this kind of performance per watt, they’d already be doing it. Years ago.
Because the latest x86 chip has to be backwards compatible with 40 years of legacy code, there are certain limitations Intel and AMD just have to live with.
Apple will likely have a hard time keeping up unless they have more up their sleeve.
It's ironic that it's Apple that will have a hard time keeping up when they've already disrupted Intel and AMD. Remember, it's not just a processor; it's a system on a chip that contains an 8-core GPU, 16-core Neural Engine for machine learning tasks and other components.
This has obviously been in the works for several years; Apple wouldn't have gone down this path if they didn't believe it would give them a sustainable competitive advantage for years to come.
Even with its ridiculous performance, the M1 is the entry-level consumer version of the M series suitable for (in Apple's lineup) entry-level consumer hardware. The M1-based MacBook Air doesn't have a fan and it's faster and more capable now than many (most?) PCs that cost more and run hotter.
One thing we know: the M2 will be faster and more capable than the M1 and won't be limited to 16 Gb. We'll probably see M2-based Macs by the summer of 2021.
We don't know what the efficiency will look like after you add all the I/O you need for a real desktop, we don't know how the memory subsystem will deal with 4 more cores that are just as fast. What's the point of having fast cores if you cannot feed them?
We don't know how lower yields for bigger chips on the newest nodes will affect their costs.
People act like Apple changed the game or is light years ahead of everyone but they literally haven't made a proper desktop-class chip yet. It's not because they don't care about higher core chips or chips with proper IO. They haven't made them because they CAN'T (yet).
So yes, Apple does have catching up to do.
That wouldn't be good enough for production hardware. The developer units had a previous iPad’s A12 SoC in an enclosure for developers to test things on. For example, the A series doesn't support virtualization because that's not required on a phone or tablet. But the M1 supports this and other laptop/desktop features. Even the A14 only has 4 cores compared to the M1's 8 cores. This isn't just a tablet SoC.
We don't know what the efficiency will look like after you add all the I/O you need for a real desktop, we don't know how the memory subsystem will deal with 4 more cores that are just as fast. What's the point of having fast cores if you cannot feed them?
Perhaps you haven't been paying attention, but the Mac mini is a real desktop. And all of the M1 Macs have blazingly fast memory access. Remember, it's an SoC—the memory is on the same die as the CPUs, GPUs and they all have equal access to it. No bus to go across. That's why it's so fast.
People act like Apple changed the game or is light years ahead of everyone but they literally haven't made a proper desktop-class chip yet.
The M1 already has features no "desktop" chip has, such as 8 instruction decoders. This may not sound like a big deal, but none of the Intel or AMD chips--not the threadripper or the Zen3—has more than 4. Why this matters:
So Intel or AMD will just add 8 instruction decoders to increase their throughput, right? Nope: As I've said in other threads on HN, there's no way Apple's first laptop/desktop chip should be competitive with AMD, but it is:> And all of the M1 Macs have blazingly fast memory access. Remember, it's an SoC—the memory is on the same die as the CPUs, GPUs and they all have equal access to it. No bus to go across.
What does this even mean? Memory works at a certain voltage, clock speed, and has a capped rate at which it can transfer data if those are kept constant. For DDR4-3200 it's generally 3,200 MTs (megatransfers/sec).
What does every core has equal access to it mean? What does soldering SDRAM change in terms how many transfers/sec it can do? What do you mean there is no bus?
> The M1 already has features no "desktop" chip has, such as 8 instruction decoders.
What makes going wide a "desktop" class feature? Going wide isn't free, there is transistor cost associated. A M1 chip has 16 billion transistors for 8 cores. AMD's EPYC Rome has 39.54 billion transistors for 64 cores. That's 8 times as many cores. And that's including ~9 billion transistors just for the I/O die.
If you look at just the chiplets, AMD's 8 Core complexes take up only 3.9 billion transistors each. And these are 8 high performance cores with SMT, even removing GPU/NPU engines from M1, AMDs design clearly has a lower transistor cost.
> The Zen3 barely beats the M1
AMD was't trying to make an entry level laptop chip, so it's not good at being a entry level laptop chip? How do you think M1 would fare if we try to compare them on datacenter CPU metrics?
I understand why everyone is excited, but they aren't looking at the whole picture.
I’ll just hit the obviously erroneous issues here.
A M1 chip has 16 billion transistors for 8 cores. AMD's EPYC Rome has 39.54 billion transistors for 64 cores.
It doesn't make sense to compare the M1 with the EPYC Rome. They couldn't more different and were developed for dramatically different use cases.
(I also don't think it's not a good look when folks feel compelled to state Apple's M1 isn't as fast as AMD's $4000 64-core processor, as if this means anything. Of course the EPYC Rome is faster--duh. Anyways…)
16 billion transistors on the M1 SoC (emphasis on system) gets you an 8-core GPU, a 16-core Neural Processor, a Digital Signal Processor, a Secure Enclave, image processing unit, video encoder/decoder and of course 8 CPU cores and 8 or 16 Gb of memory.
Instruction decoders deal with the out of order execution of the instructions; more decoders enables you to have more instructions "in flight" as they say for the execution cores to run in parallel.
Again, you should read the article for the details, but due to the legacy architecture and technical debt of a 40 year-old instruction set, not to mention with instructions ranging in size from 1 to 15 bytes, it's not easy for the decoders to juggle all of the instructions to do out of order execution of instructions. Yes, even the EPYC Rome only has 4 instruction decoders.
Meanwhile, the instructions for Apple's Firestorm cores are all the same size, making decoding the instructions much easier and faster; so much so that the M1 has 8 decoders, which is unheard of in a mainstream processor.
Here's the bottom line quote from the article: "In fact, adding more causes so many other problems that four decoders according to AMD itself is basically an upper limit for how far they can go. This is what allows the M1 Firestorm cores to essentially process twice as many instructions as AMD and Intel CPUs at the same clock frequency."
Apple and AMD solved the "get as many instructions in the pipeline as possible and process them as quickly as possible" problem two different ways.
AMD does it with threading and many cores; Apple does it by increasing the number of instruction decoders and by having 60+ Gb/s bandwidth with a 128-bit bus. Apple didn't implement SMT because they don't need it; they have no issues with keeping their instruction pipeline full.
Bottom line: Apple's approach has more headroom than AMD's approach. They can't add more instruction decoders, so they crank up the clock speed and the threads to execute more instructions per clock cycle. But the M1 is so efficient, it processes twice as many instructions at the same clock frequency.
The M1 runs at 3.2 GHz; it's got a higher ceiling because Apple can add more cores and increase the clock speed a lot before it gets anywhere near the power consumption and other physical limits that AMD is starting to run up against.
[1]: https://debugger.medium.com/why-is-apples-m1-chip-so-fast-32...
The author of your beloved article compared M1 to Zen 3, and all Zen 3 CPUs use the same 8-core chiplet. From the cheapest 6-core Ryzen 3 5600 to the 64 core EPYC. All of them are made of the the exact same chiplets, just then number of chiplets varies. If you have 2 chiplet you get 16 cores, etc.
> Yes, even the EPYC Rome only has 4 instruction decoders.
It doesn't have "4 decoders"; it's just one 4-way instruction decoder. It's not the same thing.
> Firestorm cores to essentially process twice as many instructions as AMD and Intel CPUs at the same clock frequency.
A decoder doesn't not process instructions, it only interprets them and splits it up into smaller instructions. It doesn't make sense to compare number of instructions across different ISAs anyway, ARM has smaller, simpler, and on account of that more numerous instructions. One x86-64 instruction more often than not maps to multiple ARM instructions. It just makes no sense to compare counts.
I sort of stopped reading your comment around here, sorry about that.
M1 is a clear example of something a lot better where competitors are unable to keep up or adapt.
Yet tons of people in this thread still make up (bad) reasons why Intel/AMD have nothing to worry about.
I remember similarly dumb arguments about the iPhone too in 2007. Some people are just unable to recognize major change even when it's blindingly obvious.
Status quo bias is strong.
That’s…emulation.
First line from Wikipedia[1]:
Rosetta is a dynamic binary translator developed by Apple Inc. for macOS…
Follow the link for Dynamic Binary Translator[2]:
Dynamic binary translation (DBT) looks at a short sequence of code—typically on the order of a single basic block—then translates it and caches the resulting sequence. Code is only translated as it is discovered and when possible, and branch instructions are made to point to already translated and saved code (memoization).
Dynamic binary translation differs from simple emulation (eliminating the emulator's main read-decode-execute loop—a major performance bottleneck), paying for this by large overhead during translation time. This overhead is hopefully amortized as translated code sequences are executed multiple times.
[1]: https://en.wikipedia.org/wiki/Rosetta_(software)
[2]: https://en.wikipedia.org/wiki/Binary_translation#Dynamic_bin...
Rosetta absolutely is an emulator, which has traditionally referred to taking code from one CPU architecture and making it run on another (in contrast with taking a binary of the same architecture but running it in another environment, which is often called a "compatibility layer" or in Apple's terms a "simulator", or the hardware-assisted version of this which is virtualization). There are many ways to write the part of an emulator that mimics another CPU; the simplest is an interpreter loop, but there's ahead-of-time and just-in-time code generation techniques as well. Dynamic binary translation is quite simple, as your quote mentions; Rosetta 2 instead achieves its performance through a combination of static binary translation to a form amenable to its runtime. Things that it cannot translate go through a fairly advanced optimizing JIT that goes way beyond what a dynarec would typically do.
It is not on any single Android model either. Fragmentation is a thing.
Nah, it doesn't. Source: I've got all of them
Have you ever had your laptop on a train or bus and then found out the outlets aren't powered? It's stressful!
> So far, I have experienced application crashes in Microsoft Edge, Outlook, WinZip and Logitech Camera Control. I got installation errors with Adobe Reader XI, Adobe Acrobat Reader DC, a Samsung SSD backup application, and Xbox 360 Controller for Mac.
What Mac user tries to use Edge and WinZip? A Samsung backup application? Xbox 360? Such a bizarre collection of software that snubs macOS's native browser.
https://imgur.com/a/HqxAspK
There are growing pains, and they're not all in third-party apps. (It is hilarious to me, though, that WinZip a) exists for Mac and b) is used by anyone on a Mac.)
Part of the reason for this IMHO is because your typical developer on a Mac has always used the command line tools for that sort of thing. So, Mac developers were never really motivated to build great GUI tools for certain things.
Compared to Windows, where the built in command prompt was lacking for so many years, and as a consequence 7zip/WinZIP/WinRAR gained really robust GUIs.
Every new tab incurs more RAM usage; it's not reusing. Usage keeps growing and growing until the system gets sluggish and it prompts me to kill Safari.
edit: https://twitter.com/hondanhon/status/1329132372034347009
Firefox has issues currently on M1, as well - uBlock Origin won't load, for example. Thankfully, that particular issue is fixable by disabling javascript.options.wasm and javascript.options.wasm_trustedprincipals in about:config as a temporary solution.
I fully expect these to be fixed, but it's clear it's gonna be a few weeks/months before everything runs right. If I were a non-technical user, I'd be returning it for an Intel Macbook.
The 1Blocker extension can be disabled as a temporary fix, and crucially, the ad blocking functionality does _not_ have to be disabled.
https://twitter.com/1blockerapp/status/1329744152150626304?s...
In the meantime, disabling the 1Blocker Button extension in Safari > Preferences > Extensions and restarting Safari fixes it. It won't affect how our app blocks intrusive content in Safari since 1Blocker Button only provides quick actions to the main app.
I have had Safari open for 23 hours on my M1 MBP, and it is using ~800mb of memory right now. Extensions are the most likely culprit.
Worked like a charm at least to me :-)
https://twitter.com/kb091412/status/1330470681813934086
Happy to file a report. Any special instructions to make sure it gets to you? edit: Filed one; FB8927230.
The Collections feature is also well done.
.. if you want your manager to spy on you.
MS enterprise mode can only control settings and extensions, google goes several steps further.
Says a lot.
https://www.winzip.com/mac/en/
>Xbox 360 Controller for Mac
What's sketchy about that?
Apple is pushing the "play any iOS game on the new Macs" hard. And a lot of them are supporting controllers natively on iPad and iPhone so obvious someone wants to try that out too.
I remember there used to be some special apps back in the day, to let you use a 360 controller on MacOS.
I wonder if the author knows that you don't need those apps any more?
Ditto Microsoft Edge, installed, ran a few websites - no issues, aside from a bit slow speed (although it may be just a crappy browser, never tried it on an Intel).
Perhaps with some heavier usage there are issues, but so far so good.
You seemed to be implying that this was some unique problem with OSX and that's what is holding you back from trying M1. But really, it sounds like what is holding you back is just the normal barriers of changing platforms.
Personally as a Windows user I am hoping that Windows for ARM gets supported on M1 soon, but I don't think it's so unreasonable we're not there yet, and I am not going to discount the whole platform just because of it.
What is holding me back from trying the M1, personally, is cost. I just paid $950 for a Ryzen 7 4800H with 16GB and GTX 1660 Ti. It can't hold a candle to an M1 Macbook in battery life, but I know I can run my programs and play my games and not think twice about it. My failsafe is an electrical outlet. I'm hoping I can end up trying an M1 (or successor) through an employer though! I spent the past year using a ~2014 Macbook Pro through work, and for work things, it was better than my Windows machines for working on projects where everyone else used bash on a Mac. As WSL improved, I was able to be equally productive on my Windows machines and tend to use them and not bother with the Macbook Pro.
That's all a lot of extra anecdote you didn't need, but to be sure, I'm not trying to single out OSX - the OP singled it out by saying, essentially, "you're using it wrong!"
To be a "Windows user" you have to give up the software you prefer, and use the software Windows prefers you to use. And you don't get to be unhappy if the software you prefer doesn't run OK?
And if the software I prefer runs better on macOS, I have a good reason to switch away from Windows. Just as if software you prefer runs better on Windows, you can complain about macOS (or in this case, incompatibility due to the ISA switch.)
My issue was with the comment using "mac user" as an excuse for software that isn't what the "true" mac user considers essential not working well.
In contrast, if you take the vast majority of MacOS first apps, there is no way you are ever running the vast majority of that stuff on other platforms. A wide variety of games even run on MacOS either natively or through compatibility layers like Porting Kit. In this respect out of Windows, Linux/Unix and MacOS, I think it's clear that MacOS wins hands down. It's not even close.
So given that this is pretty obvious, and I have a hard time believing you aren't perfectly well aware of this, why on earth are you trying to make this argument. Is running other platform's software really a primary criterion for you? In which case, the Mac should be right up your street.
I know its not the point, but... what's the point of using WinZip on a mac when the OS has built in unzipping?
This Mac user, having just discovered that there's a macOS version of Edge. TIL...
EDIT: well, you're off to a good start, Microsoft:
https://imgur.com/a/IWZMlMl
Instead of a browser, I guess I installed demons. I wish they'd get this shit right. MS Teams always has this notification window hiding in the background so that whenever one alt-tabs to Teams the window doesn't actually have focus and your typing goes to /dev/null. And now their browser does this "fit-and-finish" crap that few muggles will parse.
I guess I’ll have to rm it, all its junk in ~/Library and see if it gtfo’s and starts behaving again..
I use Chromium Edge. Runs great on my Intel 16" MBP.
Certainly, that second paragraph stating the M1 has many warts and is only fine for users who have a lot of money, to me, conveys a different message.
These statements are pitted against each other, even though they are both very much true:
> emulating or translating apps ... is necessarily going to be ... somewhat incompatible at best ... But Apple’s Rosetta 2 ... is a technical marvel.
He's upset about this quote from Moorhead's review:
> there were some very positive things about the new laptop. The new M1 processor is impressive, but far from perfect — it has many warts
Moorhead perhaps punctuated this poorly, but I think it's reasonable to assume that the warts refer to "the new laptop" in general, not just "the new M1 processor" in particular. Yes, his problems are generally software related (although he cited fan noise as a complaint), but that's a huge part of the product. Lots of people would love to use an M1 Mac without macOS, including one particular Finn.
Gruber can say, "There is no 'Well, here’s the downside' with regard to the state of Apple Silicon," but that remains to be seen. How much of the M1's advantage is due to a trillion dollar company buying up every last bit of TSMC's 5nm production capacity? Can they scale the unified memory architecture much beyond 16 GB? If they can, great; they've truly lapped the field.
> There is no balance, if by balance, you’re looking for a story that says any PC hardware, ARM or x86, is competitive in any way with the M1 Macs for low-energy computing.
This review wasn't about "low-energy computing." It was about using a new Mac as a replacement for an old Mac. This is a real, shipping product, not a beta. You can snicker at Moorhead's Microsoft-centric workflow and choice of apps., but that's his perspective.
Yes, it's very hard to separate process from architectural (and implementation) design. And, as an industry, we're probably sticking our fingers in our ears more than we should about what happens when process shrinks really truly stop happening. And they will. There are still a variety of options but there's very little evidence they have anything like the legs of CMOS process scaling even if we do dump a big part of the problem on developers who don't get to develop for a largely unified architecture any longer or lean as heavily on high-level abstractions.
Every other review out there covered the strong compatibility and how amazing the M1 is. I for one am glad that there exist a least a couple of reviews that focus on everything that doesn't work. Let's be honest, no one these days makes a purchasing decision based off of just reading the first review they find.
1) Graphic Design software: Works well out of box. Key Adobe apps not yet native and have some minor issues under Rosetta, but still work, are still fast and should be much faster when native.
2) Video Production: Unbelievably fast, on par with high end iMacs and Mac Pros and should get faster as all tools go native.
3) Software Developers: Super fast builds and tools. Homebrew not native yet but can still use with Rosetta version of terminal. If you need Docker it’s months away.
4) niche Users of esoteric apps ported from windows: this is the group Patrick reported in. Yea, some of these apps are months away from being usable but all have perfectly cromulent native replacements.
For certain values of "cool", for certain tasks.
iPhones and iPads sure can run quite warm when charging and doing intensive things on them.
https://news.ycombinator.com/item?id=25285489
Try to submit this same link now and that's what will happen. This feature already exists, it's just that it seems too narrow of a window if it's only a few hours long.
There's no denying the part of our brain that gets a kick out of the internet points which reflect peer approval and respect. However, I will continue to hope to value the greater good that this community creates for all of us over the individual recognition that gives me a much shorter term buzz.
Aside from the peer approval that you mentioned, I also enjoy the idea that I can somewhat steer the conversation and see what smart people have to say on the topics I’m most interested in.
The reason we allow reposts of articles that haven't had significant attention yet is that it's largely random what happens to get noticed on /newest, and we want to allow good articles multiple rolls of the dice. The priority is on the article, not who submitted it. That does make it a bit of a lottery, and it's winner-take-all. There are definitely better solutions possible, and it's on my list to work on that someday. We'll expand how karma is reward for submitting an article (HN could use some karma inflation anyhow) and find some way of grouping related submissions together.
In the meantime, the lottery does at least even out in the long run if you keep posting good submissions.
As for karma inflation, it's an interesting concept, but I don't know if I'd like that :)
And I didn’t imagine it was what drove most links to the top of the homepage. I always thought that many people submitting the same URLs was more influential.
I want long battery life, no fans and a lot of memory. Running Linux. If M1 is the catalyst to make it happen, then so be it.
The more we enable people who are unwilling to even try to learn that this particular option has moved from the Edit menu to the Format menu, or has removed the default toolbar icon (but you can add one), or has simply moved an inch to the left, the more we send the message that it's normal and expected that people shouldn't have to ever adapt and change and learn new things.
One big example of that is the removal of community captioning for YouTube videos. Call me hidebound but I am not happy at all with this, even though I don't use it very much. Captions are critical for making videos accessible to the deaf and hard of hearing communities. The removal of this feature is an ugly snub to a vulnerable group. Now they will be wholly reliant on the video creators to do the captioning themselves or otherwise be stuck with the lousy automatic captioning.
There is a balance to be struck, and between us, I think we've more or less bracketed it.
Backwards compatibility has long been a Microsoft obsession (out of necessity for the enterprise market) and an Apple allergy.
"I'm going to dump Apple over this!"
"See ya. Don't let the door hit you on the way out!."
Of course optical drives; that was a big step.
Apple is ruthless at pruning old tech.
I believe I read an article the other day that explained why the M1 was so fast and efficient and basically the design it came up with is impossible to do on Intel, for many reasons including the fact that Intel instructions are variable-length (which means they need to be completely parsed every time to find the end of the instruction) while M1 instructions are fixed-length (which means you don't need to parse it) due to the RISC design (finally a winner).
Methinks we are witnessing the death of the traditional Intel chip API as the bad design decisions made along the way come to a head.
Should make things interesting as now both Apple and Linux have a head start on ARM support.
Really the bigger issue is that the ARM chips Windows has thus far run on don't have the similar TSO opts the M1 came with.
(The instruction decoder would become more of a translator.)
The bigger problem with x86 is the implicit hard synchronization barrier after most writes. This cannot be easily fixed without introducing new instructions altogether.
AMD has been great (and improving) on performance/watt. It appears that the M1 has the upper hand, but then again Apple bought all the 5nm capacity for the year. I'm really curious how well 5nm Zen chips would compete non performance/watt - I suspect Zen 4 (5nm) will compete very favorably.
(also, even if it were about on M1, Linus said he thinks its unlikely, other qualified people disagree)
[1] https://elinux.org/images/a/ad/Arm-soc-checklist.pdf
I use my iPad Pro for just about everything except programming, so laptop selection is not as important to me as it used to be.
Give it a 800% overclock and dunk it in liquid nitrogen and then we'll talk.
The M1 is a great and well-refined design but when compared to AMD’s similarly priced 4750U in multi core benchmarks, the performance per watt is better but not by much more than the improved process would suggest. And that’s without the IPC improvements that mobile Zen 3 will bring.
And when compared to last gen 14nm Intel MacBooks, a 5nm TSMC part is going to blow the doors off just by virtue of feature size alone.
I think that could also be the reason it is getting so hyped, it is a tangible improvement over other CPUs. 50% more battery life, better performance in some case, what's the last CPU generation we saw such gains? Probably pretty far back to when Dennard sclaing was still a thing?
[1] https://en.wikipedia.org/wiki/Dennard_scaling
For example, with an improvement to the 14nm process and not a die shrink, Kaby Lake->Whiskey Lake saw low power processors double in core counts (2C/4T->4C/8T) in essentially the same power/thermal envelope. Basically every thin laptop/ultrabook family doubled its core count.
It's an overall improvement, but not as dramatic as "2x cores for 2x perf at the same power"
[1] https://www.anandtech.com/show/14514/examining-intels-ice-la...
The point of this is that the significant improvement from Kaby Lake to Whiskey Lake involved only small architectural refinement and updates to an existing process, so much larger performance/power improvements should absolutely be expected from an entirely new process plus architecture refinement.
I'm on a 3 month wait list right now to get an 8 core/16 thread AMD Thinkpad. Nobody else currently produces a sub-15W part with that much computational power.
The major difference vs the M1 is that AMD's micro-PCB design scales in core count rather than producing a single dense power-efficient die.
This is a big part of the reason that the AMD mobile chip comparisons seem to be kinda besides the point for me. For a bunch of reasons, its almost impossible to get a high quality laptop today w/ the latest and greatest AMD chips that people love to compare to the M1 chips. I know AMD makes great chips, but what does that matter if you can't get buy them in a laptop today?
FWIW - I love AMD, I have a Ryzen desktop for gaming. Its fantastic the competition they are bringing in the desktop market. But I don't want to play the game of searching far and wide for high quality laptops that have the new Ryzen chips _and_ are actually in stock somewhere in a configuration I want _and_ are from a reputable manufacturer.
I was able to pickup a M1 Air on release date w/ the upgraded specs from my local Apple store. It's a glorious machine, and I knew the build quality and hardware is going to be top notch.
OEMs still largely prefer Intel for laptops for whatever reason, which sucks, and it seriously hampers the ability for AMD to compete in this market.
OEMs might prefer Intel because, unlike AMD, Intel is actually able to supply the CPUs to go in their notebooks. A slow CPU is better than no CPU. My understanding is that AMD is mainly focused on supplying the gaming console makers, so I don't know that the situation will improve over the next few years for laptop manufactures seeking AMD chips.
I do find it amusing the continual retorts to M1 performance of "but wait until the Ryzen 4800/4900 laptops are out in force!" as if that were something that is realistic at all in the next year. AMD can't even keep up w/ demand for its graphics cards and console chips right now, unfortunately.
... until the end of 2020.
M1 has better single core perf, but the AMD 4800 edges it out in multi-core perf on Geekbench(apparently) and smokes it in multi-core Cinebench.
https://www.cpu-monkey.com/en/cpu-apple_m1-1804 https://www.cpu-monkey.com/en/cpu-amd_ryzen_7_4800u-1142
I wonder how many of the currently accepted trade-offs are actually false.
At low levels it's kind of easy to prove, like you can store either more or bigger packages in a truck but not both. But the higher you go, the farther from physics, the more difficult it becomes to know if you're making decisions on what is essentially a flaky trade-off.
The source of the confusion has been the benchmarking software. To saturate one core on an Intel processor you need to run two threads, because that's the way they are designed. So the single thread benchmarks that have been used so far have been using 50% of the capacity of an Intel CPU core and comparing it with 100% of the capacity of an M1 CPU core.
This article breaks it down fully: https://wccftech.com/why-apple-m1-single-core-comparisons-ar...
More discussion: https://linustechtips.com/topic/1276532-exclusive-why-apple-...
I guess at the very least CPU benchmarking software should have a "thread to thread" benchmarker alongside a "core to core" benchmarker, or something along those lines.
That would be in the spirit of having benchmarks indicate real world usage
SMT is designed to boost performance in multitreaded workloads.
It can be thought as multitasking for a CPU core.
Using two threads for one benchmark and use one thread for another, it is not comparing single core performance.
Because why wouldn't that CPU schedule the load on other cores?
The argument to be made is that because of the resources dedicated to SMT, single thread on AMD/Intel versus single thread on Apple is not measuring the true potential performance of the whole core. In principle, some multithreaded workload over all available threads could be a better metric for whole-processor performance.
If I had a single threaded application or a single threaded critical path of a multithreaded application.
I don't see a scenario where SMT will help
AnandTech have done a comparison here: https://www.anandtech.com/show/16261/investigating-performan...
You can see it's all within 1% with it on and off for single threaded.
At the end of the day, benchmarks are inherently only an approximate measure of how real-world code will perform. SMT, basically by definition, is rarely going to benchmark well, but is inherently going to show more of a benefit when running real-world mixed workloads.
The main purpose of single-threaded benchmarks is to approximate performance for things which are actually single-threaded.
>M1 Macs embarrass all other PCs — all Intel-based Macs, including automobile-priced Mac Pros, and every single machine running Windows or Linux. Those machines are just standing around in their underwear now because the M1 stole all their pants.
I think it's a bit unfair to lump Linux in there since it's been running on ARM and other embedded, low-power devices basically forever, but it is true that few of us run our Linux desktop on an ARM board.
But I think the overarching point is true: now desktop and laptop makers can't just pretend that having a hot and/or slow, clunky, noisy x86-based architecture is just a fact of life. I hope and expect that it'll help create a new generation of ARM-based laptops and maybe even desktops that will run cool and smooth.
I want to believe.
Not sure what shitty machine hes talking about, but my threadripper's are not impressed with some 4 core M1.
Anandtech ran the numbers a couple years ago and an EPYC processor spent around 60% of its budget on Infinity Fabric (up to 90% of used power with only a couple cores being used).
An M2 based chip might get more done, but it won't use much less power. In truth, I'd guess there's a good chance it bottlenecks on the IO and performance doesn't increase as expected (without widening IO further with all the extra power costs).
As it stands right now, I have an 18 month old 27" iMac with 64GB ram and an AMD gpu - and an 8GB M1. And I'm preferring the M1 for almost any task that doesn't demand the real estate.
It will not, my TR has 64 cores 128 Thread and you know whats the best thing? 128GB Ram...and imagine that..they aren't even soldered...and i even can exchange my GPU..or slap in 2 others.
And if apple ever brings out a M1 Workstation, i can probably buy 3 other Server/Workstations with that money.
Firestorm Cores use less than 4 watts each. A 16 firestorm core Mx will produce roughly the same multicore performance as your TR, with a less than 70 watt TDP.
And a $65 M1 dusts a 4800h that costs over 4x as much and using 1/3 the power. Top end ThreadRippers cost $2,000 to $3,000 each. You can bet an Apple Mx ThreadRipper killer isn’t going to cost more than $500 given it only needs 4x the cores of the $65 M1.
I suspect we'll never know. Apple have shown little interest in anything other than prosumer workloads for a long time. I doubt they'll be making an M1 server platform anytime soon.
It's possible Apple just doesn't think the investment in a 30 core CPU is worthwhile and stops around 16 cores, which would still be higher than any i7 or any i9 ever made.
https://images.anandtech.com/doci/13124/IF%20Power%20EPYC.pn...
Given that scaling, you're looking at 140w between 64 cores which is right at 2w per core. The all-core turbo for TR is 3GHz. If we assume linear scaling on M1 cores, if you dropped them in to replace the x86 cores and the had to reduce from 3.2GHz at 4w down to 2w per core, the all-core turbo would be a mere 1.6GHz AND about a third less cache per core.
TSMC claims 30% less power going from 7nm to 5nm. That means those x86 cores power usage drops from 2w to 1.4w. With the same linear scaling and 4w at 3.2GHz, the M1 design is now operating at 1.1GHz competing with a part that is still 3GHz and the M1 still has only 3mb of cache per core instead of 4.5mb.
Now, I'm betting that actual power usage is 3w which bumps frequency up to 1.5GHz. Still half the frequency, but possibly close enough in performance to Zen 3 at that frequency.
EDIT: It's apparently 3.45w during cinebench according to these benchmarks.
https://twitter.com/panzer/status/1328715510100344833
I'm betting that if they didn't bottleneck on the IO that the M1 would perhaps be faster, but at that scale, interconnects are a much, much larger issue than the cores themselves.
Zen 3 chiplet is 4.15B transistors. The consumer IO die is 2.09B transistors. The Threadripper IO dies is 8.34B transistors. M1 is 16B transistors. A14 is 11.6B transistors. That difference is mostly 2 firestorm cores/cache, 4 GPU cores, another DDR4 controller, and some Thunderbolt/PCIe.
Taking a ruler to a die shot and multiplying that area by transistor density (16B/119mm^2) gives us somewhere around 500M transistors per core plus 3mb of cache. Adjusting out to 4.5MB per core to match Zen 3 adds another 75-100M transistors.
Zen 3 gives 518M transistors per core, but that includes interconnects. Taking a ruler to the core reveals about 14% of the die is interconnect, SMU, and wafer tests (about 580M transistors). Subtracting that from 4.15B and dividing into even core/cache gives 446M transistors per core.
We'll assume an identical IO cost of 8.5B transistors and 150M transistors of interconnect per core. (64 * (600 + 150)) + 8500 is 56.5B transistors and 420mm^2 total area. A zen 3 chip with similar densities would be 310mm^2.
No matter how you slice and dice that, Apple's chip is going to be significantly more expensive than AMD's chip. Thinking anything different is a pipe dream.
EDIT: also note that formal verification adds a lot to chip pricing.
This is just not true. A mini going from idle to a full single-core load is a power increase of 6W. A heavy multicore load hits over 25W.
> A 16 firestorm core Mx will produce roughly the same multicore performance as your TR
With what fantasy math? 30% faster single core performance would be 16 * 1.3 = ~21 equivalent TR cores. It'd be less than half the performance of a 64 core ThreadRipper.
> And a $65 M1 dusts a 4800h that costs over 4x as much and using 1/3 the power
https://images.anandtech.com/graphs/graph16252/119365.png
M1's single-core performance is very strong. But not enough to overcome a ~2x or greater core deficit, either, not at all.
I can't live without 64GB in my desktop and would love to have more, so the new Macs are not for me.
1: https://d3nevzfk7ii3be.cloudfront.net/igi/ZRQGFteQwoIVFbNn (from https://www.ifixit.com/News/46884/m1-macbook-teardowns-somet... )
Hardware companies don't charge cost + some reasonable markup, they charge whatever the market is willing to pay for it. I highly doubt Apple would release some hypothetical "threadripper killer" at a price any lower than they needed to to compete.
Lets assume that all the other parts, high speed RAM, SSD, power supplies, cooling, motherboard, etc cost another $500, and that operations, assembly, marketing etc average 20% and the manufacturer aims for a 5% margin. So they sell their low end Threadripper PC for $3300, and their high end Threadripper for $4,700.
Then Apple comes along and makes it's own high core count versions of the M1 (lets call them the M16 and M24) for $1,000 and $1,500 with similar performance to the Threadrippers. When Apple builds high end Mac Pros out of the, they also spend $500 for other parts, giving total parts cost of $1,500 and $2,000, it marks them up 50% for their typical margins, and sell them for $3,000 and $4,000 and have the same performance as the ThreadRippers from the lowest margin manufacturers at for 10-15% less.
And on top of that, the M16 and M24 use less power and generate quite a bit less heat than the Threadrippers. They don't need any fancy cooling systems. That can lead to smaller packaging and even more cost savings.
Obviously when Apples 16 and 24 Firestorm core Mx chips come out newer Threadrippers will have evolved to smaller process as well, and that will help increase their performance, while reducing cost and heat. But there is little doubt that Apple's Mac Pros will be much more competitive price and performance-wise at that time, and if they keep leveraging TSMC first there is a significant risk that they actually take some of their laptop cost/performance advantage up to Threadripper levels.
Based on the MacBook Air/ Mac mini/ MacBook Pro pricing, I don't think this is the case.
Apple has delivered a significantly faster MacBook Air for the same price as the previous generation. Not only did they increase the CPU speed, they essentially doubled the performance of their the SSD while maintaining their base price.
Apple dropped the price of the base mini by $100 while delivering a significantly faster Mac mini.
I suspect whatever Apple delivers for their top end machines will be equal or lower price to their current Intel machines.
The apple also has twice as many memory channels (8 vs 4) and runs the memory significantly faster (4266 vs 3200).
The m1 also has about half the latency to main memory.
So for highly random workloads that aren't cache friendly I'd expect the M1 to win.
You're quite right, your machine and in fact everyone else's are still just as good as they were a few weeks ago. I got an Intel 5K iMac in the summer and I'm still very happy with it. I need to be able to run windows and Linux in VMs so M1 isn't for me right now. It's no for you either.
Having said that, your manufacturer of your Threadripper is now on notice. Apple isn't done with the M1, more powerful desktop chips are on their way over the next few years and there's a very good chance they'll do exactly what they just did to Intel and AMDs mobile chipsets to the high end stuff. Apple are coming for them, and not being impressed now isn't going to serve them well down the line.
Going back to versatility, there's no way that AWS or Azure is going to move to Apple silicon for their infrastructure, so there's still a massive market for AMD or other Arm manufacturers to target. For the consumer side unless Apple releases a chip that results in them making a sub-$800 laptop, they're just simply not going to capture a large market. With the rise of devices like Chromebooks in education and home there's still something to be said for Apple devices just being too expensive.
In general, Apple devices also leave a LOT to be desired for any sort of large scale enterprise organization. Unless they go back to their 90's and early 00's mindset, I just can't see this changing.
It's certainly exciting to see the CPU market being upended by Apple and AMD, as well as the potential future for a large architecture shift, but I don't think there's any real threat to market shares as a whole.
AWS already have a series of AArch64 instance-types which they will most likely continue to improve.
This is exactly what Apple did with the iPad, nowadays you can get a very decently specced 10.2” iPad with their 2018 CPU for $309. That’s astonishing, but it’s a great way for Apple to get extended value out of their silicon.
But claiming that what’s available today from Apple is of great interest to the high performance market is jumping the gun.
There’s also the fact that cost (including peripherals and support) and software availability, not performance and efficiency, are what guide most PC purchases. The M1’s advantages are far less pronounced in those areas.
I think that misses the point that most users (and most software developers in particular) use laptops. The M1 SOC pretty well kills everything CPU-wise, and does graphics as well as a GT 1050 discrete GPU. It also throws in AI silicon and RAM, and runs all of it in a 20W (or so) power budget.
Single-thread performance is quite comparable to a Threadripper.
It'll be interesting to see how things look when M1 style silicon is introduced for the 16" Macbook Pro (guessing 8 performance cores, 32 GB RAM), and the Mac Pro (guessing up to 32 performance cores, and at least 64 GB RAM used as cache for a potentially much larger main memory).
> There’s also the fact that cost (including peripherals and support) and software availability, not performance and efficiency, are what guide most PC purchases. The M1’s advantages are far less pronounced in those areas.
Actually as to cost, it's highly competitive in the premium laptop segment (like Dell XPS for instance).
Software availability is good, since it'll run almost all Intel Mac software, all M1 Mac software (soon to be ~everything), and all ARM-compatible Linux software (ARM Linux runs in a VM).
The only thing missing is Windows, and that may well be temporary.
I'd say if you're doing web, Mac or Linux software development the Apple Silicon Macs are the wave of the future! :-)
This move is very Nintendo-like. Gimmick then don't even attempt to compete with the big dogs.
The fanboys have their meat to chew on "it's not hot". And they will repeat it on command as Apple falls behind to a swath of competition. "It might be slow, but at least it doesn't warm my lap."
Apple just dropped hardware which crushes every comparable competitor.
But to you, Nintendo is not competing by "making more money", and Apple is "falling behind" by ... jumping years ahead? This reads like personal axe-grinding that is detached from any objective link to the market.
https://www.cnbc.com/2020/10/28/sonys-second-quarter-profit-...
https://www.cnbc.com/2020/11/05/nintendo-raises-switch-forec...
Well, they just pantsed Microsoft at their own ARM laptop game (#2 most valuable) and the #1 big dog is, er, Saudi Aramco.
To me, this means that when Apple consciously goes after the high end desktop market, they are likely to overshoot in exactly the same way M1 overshot the mid range mobile space. That could put Apple’s high end iMac chips in Mac Pro territory anyway.
The M1 has 8 CPU cores, 8 GPU cores and 16 Neural Engine cores. It also runs at 10 watts and maxes out at 24 watts. A Threadripper starts at over 100 watts…
Not sure if it maxes out at 24 watts. Someone measured wall power of a Mac Mini hitting 21 watts at full load, but that's including power supply inefficiency, all of the supporting circuitry and IO, a fan, flash storage, etc.
I think you're underselling the small cores a bit. Their microarchitecture is roughly comparable to a Cortex-A75, which was the high-performance cores in Snapdragon 845 phones like the Samsung Galaxy S9. But Apple clocks them about 26% slower than the SD845 did, so maybe we should be comparing against Galaxy S8 instead: flagship Android performance from 3 years ago. That's capable of a lot more than just casual background processing. Samsung introduced their DeX docking station for smartphone-powered desktop usage back with the S8 generation, when their phones only had as much computational power as the Apple M1's smaller cores.
The Firestorm cores in the M1 are 4.5X faster at SPECfp than the Icestorm cores. The Firestorm cores in the M1 are 3.1X faster at SPECint than the Icestorm cores.
In my world, the 3.1 - 4.5X faster sibling makes them very low performance, relatively. In a fully working together scenario, the "8" core M1 is equal in performance to a conceptual 4.5 core Firestorm part (see the Geekbench scaling, for instance). This, again, has positively nothing to do with any other core or part.
Bizarre that anyone thinks this is somehow negative or dismissive of the M1. It is an extraordinary chip. But they intentionally, purposefully put 4 very low performance -- compared to Firestorm -- cores because it matches the computing model of users average workload.
Power users understand today that max MHZ doesn't even matter anymore, for various reasons.
I think power consumption (even in desktops, servers) will be a large factor / marketing number.
I think the big tech sites will continue to hammer home some type of "CPU work / watt" type metric, which is roughly like "time to complete some standardized workload vs. power used to do it"
People buying a threadripper today don't care about it.
What matters to them it's the performance, my gf is a 3d artist, she doesn't care if some cpu consumes more or less power than another (when I say she doesn't care I mean she doesn't even know that CPU have very different power requirements), she needs the fastest gear she can buy on her budget.
Apple M1s are not it.
There are many other that do not care because the hardware stays at the office, where someone else pays the bills and they too don't care because the electricity bill is the last of their problems, fixing it buying more efficient computing devices would mean spending a lot of money in advance to replace at least half of the stock.
How many more months of electricity could that money pay?
A lot.
People buying Apple for its M1 low power consumption are a niche inside a niche.
So I think the reasoning stands: Apple M1s are not a real threat for its competitors because the market that really cares about its strengths is smaller than the one that doesn't and Apple will keep to largely miss the second one.
Also, the fact that the M1 is even being mentioned along with a desktop powerhouse like the threadripper says it all. The M1 is the lowest end mobile Apple Silicon. We know higher end mobile chips are coming along with desktop ones. The M1 has put all the chip manufacturers on notice, much the same the way the iPhone did (to the point RIM thought Apple was lying about the iPhone because they thought the device was impossible).
Now if we look at what people are buying and ask ourselves 'are these the people who care about TDP?' the answer is simply 'no'
The majority of laptops sold are sub 700, as the previous post said, low budget, low power and in education, markets that Apple has never even considered (except education, but it's mainly college students in USA)
Not saying M1 aren't good, but that they won't shake the market as NVidia buying ARM could
P.s. M1 is a Zen 3 competitor, threadripper has been mentioned for the exact opposite reason: because that's the state of the art for people looking at performances in personal computing (while in HPC ARM is prominent)
This is a joke. The M1s performance is no where near the threadripper. Even on per-thread performance, new AMD cpus perform better than the M1... and they have many more cores. The M1 can only win these comparisons when we start adding all kinds of conditions, like power draw, heat, price point, etc. They need to add those conditions to redirect the conversation, because on raw performance threadripper destroys the M1.
The article made the mistake of leaving out those conditions to make it seem even more impressive than it is... but hyperbole and fantasy will always sound better than reality. Reality requires those conditions. That doesn't detract from the m1's impressiveness.. but it's foolish to forget that and go off and start making ridiculous claims like the article has done.
I don't do the most intensive tasks on thin, battery-powered computers for a reason. I try to do those on the box that's always plugged into mains power with heatsinks and fans as needed, plenty of ports to plug things in, and a big monitor. A more performant portable computer is awesome, but until I can spend a similar amount to make my workstation smaller without losing capability, I'll likely keep using it for heavy lifting alongside a cheaper, less powerful laptop or tablet for lighter work or remoting in.
> Apple M1s are not it.
If a person had a $1000 budget, they could by an M1 equipped laptop from Apple. Is there something significantly faster at that price?
I just looked at Dell and an XPS desktop with 16 GB of RAM starts at $650. You aren't going to be able to add two GPUs and a monitor for $350.
At HP you can get to 32 GB for $650 but again you still need to buy GPUs and a monitor.
A lot of studios sell their old GPU (as in a generation ago) for cheap
If you stay out of vertical integrated supply chains you can buy the best your budget allows
If you are saying that you cannot assemble a laptop of the same quality of an Apple one, I agree
If you are saying that Apple laptops are what people want because they consume 20 watt full load, I disagree
The market of the good enough products is always gonna be larger than 'the best money can buy'
Apple will never make 'good enough products' hence they are not competing with the bulk of what is being sold right now
It could change in the future, but not in the immediate future IMO
Not all RAM is equivalent. If one wants 64GB of the kind of RAM M1 uses, which is very high performance, it will cost more than 300 euros.
I guarantee that 64GB does not cost anything near EUR300.
You might be thinking of HBM[2] which has a wider I/O path and costs more.
showed https://www.cpu-monkey.com/en/cpu-apple_m1-1804
which determined the M1's memory is LPDDR4X-4266 or LPDDR5-5500. If those memories are not high performance, what is?
Who cares if an M1 consumes less energy than a candle if I can buy 64GB of DDR4 3600 for 250 bucks and render the VFX for a 2 hours movie in 4k?
Another 300 bucks buy me a second GPU
When I deliver the job I put aside another 300 bucks and buy a third GPU
Or a better CPU
vertical products are an absolute waste of money when you chase the last bit of performance to save time (for you and your clients) and don't have the budget of Elon Musk
The M1 changes nothing in that space
Which is also a very lucrative space where every hour saved is an hour billed doing a new job instead of waiting to finish the last one to get paid
You can't mount your old gear on a rack and use it as a rendering node, plus you're paying for things you don't need: design, thermal constraints, a very expensive panel (a very good one, but still attached to the laptop body, and small)
So no, M1 is not comparable to a Threadripper, it's not even close, even if it consumes a lot more energy
When I'll see the same performances and freedom to upgrade in 20W chips, I will be the first one to buy them!
https://www.newegg.com/corsair-64gb-288-pin-ddr4-sdram/p/N82...
Then there's the 92% (actually 92.4%) of the remaining market that is not using an Apple computer that will keep buying non Apple hardware
Even if Apple doubled their market share, it would still be 15% Vs 85%
How is it possible that on HN people don't realise that 90 is much bigger than 10 and it's not a new laptop that will overturn the situation in a month is beyond me
Ummm... ok. But my comment was not all RAM is equivalent.
I guess you don't drive a Ferrari or a Murciélago
And does it really matter to have a faster car if you can't use it to go camping with your family because space is limited?
That's what an Apple gives you, but it's not even a Ferrari, it's more like an Alfa Duetto
It's not expensive if you compare it to similar offers in the same category with the same constraints (which are artificially imposed on Macs like there's no other way to use a computer...)
But if you compare it to the vast amount of better configurations that the same money can buy, it is not
Yeah… no, those days are over. The reviews clearly show the M1 Macs, including the MacBook Pro outperform most "desktops" at graphics-intensive tasks.
>So no, M1 is not comparable to a Threadripper, it's not even close, even if it consumes a lot more energy
Um… nobody is comparing an M1 Mac to a processor that often costs more than either the M1 Mac mini or MacBook Pro. However, the general consensus is the M1 outperforms PCs with mid to high-end GPUs and CPUs from Intel and AMD. Threadripper is a high-end, purpose build chip that can cost more than complete systems from most other companies, including Apple. However, it's at a cost of power consumption, special cooling in some cases, etc.
>Who cares if an M1 consumes less energy than a candle if I can buy 64GB of DDR4 3600 for 250 bucks and render the VFX for a 2 hours movie in 4k. Another 300 bucks buy me a second GPU
The MacBook Pro has faster LPDDR4X-4266 RAM on a 128-bit wide memory bus. The memory bandwidth maxes out at over 60 GB/s. And because the RAM, CPU and GPU (and all of the other units in the SoC) are all in the same die, memory is extremely fast.
From AnandTech; emphasis mine [1]: "A single Firestorm achieves memory reads up to around 58GB/s, with memory writes coming in at 33-36GB/s. Most importantly, memory copies land in at 60 to 62GB/s depending if you’re using scalar or vector instructions. The fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before."
It can easily render a 2-hour 4k video unplugged in the background while you're doing other stuff. And when you're done, you’ll still have enough battery to last you until the next day if necessary. According to the AnandTech review [1], it blows away all other integrated GPUs and is even faster than several dedicated GPUs. That's not nothing; and these machines do it for less money.
>vertical products are an absolute waste of money when you chase the last bit of performance to save time (for you and your clients) and don't have the budget of Elon Musk
>The M1 changes nothing in that space
This is not correct… seeing should be believing.
Here's a video of 4k, 6k and 8k RED RAW files being rendered on an M1 Mac with 8 Gb of RAM, using DaVinci Resolve 17 [2]. Spoiler: while the 8k RAW file stuttered a little, once the preview resolution was reduced to only 4k, the playback was smooooth.
[1]: https://www.anandtech.com/show/16252/mac-mini-apple-m1-teste...
[2]: https://www.youtube.com/watch?v=HxH3RabNWfE
Don't get me wrong - this is still very impressive with a ~20w combined! power draw under full load, but it definitely doesn't beat mid - high desktop GPUs.
(This is largely irrelevant for video encoding/decoding though as you can see - as that's mostly done either on the CPU or dedicated silicon living in either the CPU or the GPU that's separate from the main graphics processing cores.)
[0] https://www.macrumors.com/2020/11/16/m1-beats-geforce-gtx-10...
You could build a complete desktop system including a GPU that's more powerful than the one in the M1 for ~$1000, but certainly not a 3080. They're very expensive, and nobody has any in stock anyway.
An RX 580 or 1660 would probably be the right GPU with that budget. (Although you could go with something more powerful and skimp out on CPU and ram if you only cared about gaming performance).
- yes, you totally can. The best thing is that with a 1k.entry level you can start working on real-life projects that have deadlines and start earning money that will let you upgrade your gear to the level you actually need, without having to buy an entire new machine. The old components can serve as spare parts or to build a second node. You don't waste a single penny on things you don't need.
Even though, it's true, you can't brag with friends that it absorbs only 20 watts full load and the heat of the aluminium body is actually pleasant
It's a big sacrifice, I understand it.
They don't!
Cut the BS
> Here's a video of 4k, 6k and 8k RED RAW files being rendered on an M1 Mac
Blablablabla
That's not rendering
AnandTech disagrees with you: https://www.anandtech.com/show/16252/mac-mini-apple-m1-teste...
Besides the additional cores on the part of the CPUs and GPU, one main performance factor of the M1 that differs from the A14 is the fact that’s it’s running on a 128-bit memory bus rather than the mobile 64-bit bus. Across 8x 16-bit memory channels and at LPDDR4X-4266-class memory, this means the M1 hits a peak of 68.25GB/s memory bandwidth.
Later in the article:
Most importantly, memory copies land in at 60 to 62GB/s depending if you’re using scalar or vector instructions. The fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before.
Dual channel DDR3L or DDR4L also has a 128 bit bus. 4200MHz DDR4 is clocked on the high side for most laptops, sure, but it's hardly unusual.
Run the numbers and you get the exact same throughput figure as for M1, which isn't surprising, because we're just taking width * rate = throughput.
So I'll repeat my assertion, downvotes be damned: the memory on the M1 is not special. The packaging and interconnect is interesting. It might reduce latency a little; it probably reduces power consumption a lot. But there's nothing special about it. The computer you're on right now probably has the same memory subsystem with different packaging.
That’s where they started, but their conclusion was beyond that.
Did you miss the part where they said the fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before?
This isn’t only about A14 vs M1.
It’s not that LPDDR4X-4266-class memory is special; it’s been around for a while. What is special is that the RAM is part of SoC and due to the unified memory model, the CPU, GPU, Neural Engine and the other units have very fast access to the memory.
This is common for tablets and smartphones; it’s not common for general purpose laptops and desktops. And while Intel and AMD have added more functionality to their processors, they don’t have everything that’s part of the M1 system on a chip:
* Image Processing Unit (ISP) * Digital Signal Processor (DSP) * 16 core Neural Processing Unit (NPU) * Video encoder/decoder * Secure Enclave
There’s no other desktop such as the M1 Mac mini that combines all of these features with this level of performance at the price point of $699.
That is special.
I don't think that's notable, sorry. I would expect that of any modern CPU.
> it’s not common for general purpose laptops and desktops
Well, yeah, because "memory on package" has major disadvantages. You (laptop/desktop manufacturer) are making minor gains in performance and power and need to buy a CPU which doesn't exist. Apple can do it, but they were already doing it for iPhone, and they must do it for iPhone to meet space constraints.
I think unified memory is the right way to go, long term, and that's a meaningful improvement. But as you point out, there is plenty of prior work there.
> they don’t have everything that’s part of the M1 system on a chip
They actually do! The 'CPU' part of an Intel CPU is vanishingly small these days. Most area is taken up with cache, GPU and hardware accelerators, such as... hardware video encode and decode, image processing, security and NN acceleration.
Most high-end Android cellphone SoCs have the same blocks. NVIDIA's SoCs have been shipping the same hardware blocks, with the same unified memory architecture, for at least four years. They all boot Ubuntu and give a desktop-like experience on a modern ARM ISA.
> There’s no other desktop ... at the price point of $699
Literally every modern Intel desktop does this.
The best that money can buy changes when you put the actual budget. Today, Apple may be the best that $1000 can buy.
Em NO, the M1's are not comparable with a workstation class laptops (not cpu wise nor ram...and especially not GPU wise)
Got it, watching Netflix in Chrome without the laptop getting hot is "a niche inside a niche" while GPU intensive 3d graphics workflows are mainstream.
Don't need no M1
AMD and Nvidia have sold more GPUs than Apple laptops by a large margin
Well then i have to say Apple achieved what my 7 year old nexus 5 already could...bravo.
Of course I'm sure you can find laptops with terrible cooling design that would get uncomfortably hot watching Netflix, but there's plenty of thin and light laptops available for < $1000 that would do just fine too.
https://www.extremetech.com/extreme/311995-japans-new-arm-ba...
I think his point about moving the bounds of the power (and heat) / performance tradeoff is valid.
The hardest programs to run will not benefit here.
I don't need chrome to open up 0.1s faster.
Gruber greatly exaggerates here. Particularly mentioning the Mac Pro in the same paragraph.
Apple's M1 isn't competing with higher end Intel/ AMD desktop class CPUs with external GPUs. The M1 is a laptop CPU, designed for good performance with good thermals and excellent battery life. It will be interesting to see what they come up with over the next couple years to replace the CPU in the iMac and Mac Pro. But that is next year.
If we can end up with superfast, supercool ARM-based Linux laptops, that'd be a huge deal :)
While I dislike apple and have never owned any of their devices I can't help but feel thankful to them because now that we know this is possible soon others will re-create their success and we will hopefully have the same perks on a more open platform
Apple showed Debian running in a VM on an M1 prototype Mac nearly 6 months ago.
I get that it won't be acceptable for some Linux diehards but the superfast, supercool ARM laptop running Linux will soon be an M1 MacBook Pro.
...except I cannot afford a macbook. The one I'm using now was paid for by my employer.
More generally, I'd love to see fast ARM platforms not built by Apple.
And even the build quality... a friend who's an Apple fan used to tell me "they are better quality, they have less problems" yet his macbooks seem to break down at about the same rate as my Dell, HP or Lenovo laptops, so...
I've decided macbooks look good, have good quality screens and the best touchpads. I'm unwilling to pay a premium for that, and I don't even like OS X/macOS...
> I'm afraid that when/if fast ARM alternatives arrive, they will probably cost 80-100 % of what the MacBooks cost.
Possibly! I wouldn't buy them when they are the latest new thing, either.
The macbook I was given costs, after customs, taxes and everything, $4000. Not $1000. Are there cheaper macbooks? Sure. Why would I want them, though?
First, this discussion is about the new MacBook Air that costs $1000 US which is why I mentioned that price point. I suppose tariffs and such would increase that substantially, so that's a fair point. My bad. That said, I believe there was discussion about the hypothetical situation where you did want a Mac.
More to the point, I checked your post history before I commented, and I gathered you are a developer, somewhere in South America (?). I suspect that the code I write is probably no better than the code you write. But your pay is presumably quite a bit less than mine. My point was that it's unfortunate that like work doesn't yield like rewards.
Now here I am going to make some assumptions... In all likelihood, people are amassing wealth off of your work in the same way they profit from mine. But due to local wage norms, they can pay you less and reap higher margins. They get to upgrade the yacht, but you might not even have enough free time time to use a boat even if you could afford one.
It's possible that I extrapolated too much and that your situation is nothing like what I have inferred. Sorry in advance if that is the case.
You know, the M1 Mac mini is $699…
And since you can attach up to 6 displays to it (or the MacBook Pro), you'd never have to look at macOS if really didn't want to. ;-)
https://www.macrumors.com/2020/11/24/m1-macs-able-to-run-six...
Is it actually possible to replace macOS with Linux though?
The main issue is driver support, since graphics, networking, etc. is all Apple proprietary hardware.
I have wanted a powerful "enough" fanless laptop for many years. I got excited about 4 years ago with Intel's short-lived M line and I bought an Asus ux305. It served me well for front end web dev (had to use a cooler stand under sustained high load though). Great Linux support and battery life (after installing tlp). I remember feeling a bit of awe that it was reasonably snappy with no fan, but it was not quite there. I can only imagine the feeling is an order of magnitude greater with the M1.
I don't own any Apple products, but I still hope that raising the bar will benefit me later on. I'll be pretty pumped when this level of hardware can run a mainline kernel.
This actually worries me.
Lately, anytime I look outside the Apple ecosystem, I wonder to myself what all these companies are doing... sure, I could see AMD closing the gap, but Apple's made additional choices for integration beyond what AMD would do alone.
I guess my summary would be: other companies trying to "match" Apple hasn't really gone well in the past decade, short of a few select standouts.
My main use machine now and for last couple years has been a Google Pixelbook. Lightweight (2.4 lbs) and fanless, solid 8+ hours on a charge, great for general use on chromeos, okay for Linux with built-in Crostini container.
I've had an iPad, but have never had any other Apple product. I've been pretty much anti-Apple since 1979, when as a 15-year-old I bought a Radio Shack TRS-80 over the much more expensive and glitzier Apple IIe.
I now have a Macbook Air M1 on order. We'll see how it goes. I'm sensitive to weight, prefer very light laptops. At 2.9 lbs it will be a noticeable step back from the ux305 (2.6 lbs) and Pixelbook (2.4 lbs). If I end up not bonding with the Macbook I plan to sell it; Apple's generally high resale value is a factor in giving the M1 a try.
but growing more today than it was a few years ago because of the prevalence of Chromebooks in education. Many of which run ARM CPUs!
ARM based Linux server CPUs? Very likely. Low end/ Chromebook style laptop ARM CPUs, Already exist.
Performant ARM CPUs designed for laptops? I expect you just gotta hope Windows on ARM takes off or start porting Linux to M1 Macs.
https://www.techradar.com/news/you-can-now-run-linux-and-win...
What makes you think Windows won't take off on ARM?
In any case, if it can be done -- i.e. if there aren't locked down specs -- you can bet Linux will be ported to M1 Macs.
I have no idea if it will or not. Microsoft's current strategy on ARM seems pretty unlikely to succeed though.
> In any case, if it can be done -- i.e. if there aren't locked down specs -- you can bet Linux will be ported to M1 Macs.
I'm confident (and hopeful) we will see an M series Linux port. How good it will be without support from Apple is questionable. The current state of Linux on Mac doesn't exactly inspire hope.
I'm not familiar with Microsoft's current ARM strategy, but they can always pivot of it turns out it's really the future. I must reluctantly admit Microsoft has been capable of some pretty surprising and welcome changes in its history.
I also wonder about the driver situation. Intel hardware has excellent support on Linux, everything works out of the box. Apple probably has zero interest in contributing drivers to the Linux kernel.
I can dream of an even better one, though!
My key assumption is that there's some pareto principle at play--80% of the M1's performance gains over the next best generic ARM chip are due to a handful of key innovations while the remaining 20% come from a long tail of incremental improvements. This would mean that generic manufacturers could copy those handful of key innovations (modulo IP rights, which might be another fatal flaw in my speculation) and recoup 80% of that performance gap in just a few years. If that assumption is bad and the performance is entirely an accumulation of lots of minor innovations over time, then I suspect it will take chip manufacturers quite a lot longer to catch up.
Very handwaivy but the devil is in the details of that "all they have to do". Copying success is not that straight forward or else AMD would have had a response to Conroe in 2007, Intel for Zen in 2018, and Qualcomm for A-chips around 2014.
When iPhone came out in 2007, Android did not even exist. [1]
[1] https://en.wikipedia.org/wiki/Android_(operating_system)
> An early prototype had a close resemblance to a BlackBerry phone, with no touchscreen and a physical QWERTY keyboard, but the arrival of 2007's Apple iPhone meant that Android "had to go back to the drawing board".
I guess Android, or what morphed into Android, was probably being developed, in a pre-beta state, when iPhone was announced in January 2007. Okay. It sounded to me like someone was saying that Android was released on Blackberry-like phones at the time iPhone came out. That was only thing i was objecting to. Sorry, a small point I almost didn't even bother to make, and probably not helpful to anybody, anyway.
[1] https://en.wikipedia.org/wiki/Android_version_history
I'll be interested to see if Apple can keep this up with 128GiB of RAM and a beefy GPU.
So far to me M1 is great, ARM is great, but comparing a tiny amount of RAM soldered onto the CPU compared to flexible and huge RAM potential devices seems dishonest. Unless Apple manages to squeeze 128GiB of RAM on their chip, at least.
I want to see what my Blender workstation looks like on Apple's Architecture. Where i purpose built a huge CPU, GPU, and RAM for Blender rendering and large scenes. I suspect Apple will start seeing huge falloffs in the SoC design under larger workloads.
Apple is dominant in making phones though. And this Laptop is designed like a Phone. So yea, it's amazing for that.
The magnitude of extra ROB entries, having an 8x wide instruction decoder, that massive L1I cache, and such fat cores with lots of ALU and other op units while simultaneously keeping over a 3GHz clock is a phenomenal architectural improvement overall.
Apple may not need to add more RAM locally. Other silicon designers like IBM with POWER10 are moving to serial memory with massive "L4" caches, seen previously with their Centaur memory controllers. I can see Apple using OMI or a similar serial memory bus to add extra memory, using 16/32/etc GB of PoP RAM for caching reads and writes to slower off-chip memory. I expect Intel and AMD will eventually move to a similar memory model, at least in server grade hardware.
[0] https://www.anandtech.com/show/16226/apple-silicon-m1-a14-de...
Sidenote, even if Apple rolled out 128GiB tomorrow i probably couldn't afford it. As the OP says:
> including automobile-priced Mac Pros
That's rough.
Side note, while Apple charges ridiculous prices for their upgrades, Mac Pros are actually some of the more ""reasonably priced"" computers comparatively. I've looked at HP and Lenovo's equiv. workstations, and you pretty much get screwed by everyone at that scale.
My desktop can take up to dual 128 GB sticks but one stick costs over $1,000, so I'm stuck with my 32 GB stick :(
... OTOH, it's pretty insane you can get 128GB of RAM for $1,000 nowadays. I remember when 8GB was a lot and cost several hundred. Time and miniaturization go on I guess...
It's just very close to the CPU, on the same package. It's not on the same die or anything. There are laptops that already have a shorter physical path to RAM than the M1 MacBooks, stacking the RAM right next to the CPU is common.
So it's just a matter of time before >16GB is available with an M1. These are just entry level machines in the first round.
Yea, i was just simplifying. Though in SoC i think that's a perfectly fine simplification lol.
> So it's just a matter of time before >16GB is available with an M1. These are just entry level machines in the first round.
Think there will be > 32GiB?
I mean, what would the market be for a super-fast ARM laptop/desktop machine? It would run Windows RT, I guess?
How much better than Intel/AMD would the ARM machine need to be to justify buying it and running that? It seems like the M1's "magic" is not just the chip, but very much also the Rosetta thing.
IMO, one of the key innovations Apple made in the M1 is a modification to the memory model so that it is much more efficient to translate x86 code into ARM code.
There's lots of credible ARM work in both the mobile and the server space. Generally, a lot of it would work OK in the PC space, but for the lack of software support. Apple has the ability to leapfrog that particular problem.
Still, Apple seems to have done pretty well compared to both the other mobile and server ARM vendors with their PC offering.
I do agree that Apple's push to ARM for laptops will have a dramatic impact on the server space in a year or two. Intel is in for a world of hurt unless they can pull a rabbit out of their hat.
On October 1, 2019, rumors emerged that Samsung had laid off their custom CPU teams at SARC.[21][22][23] On November 1, 2019, Samsung filed a WARN letter with the Texas Workforce Commission, notifying of upcoming layoffs of their SARC CPU team and termination of their custom CPU development.[24] SARC and ACL will still continue development of custom SoC, AI, and GPU.[25]
The Raspberry Pi 4 / Raspberry Pi 400 are halfway decent to use. The Lenovo Chromebook Duet is probably the first ARM-based Chromebook where I haven't felt constrained to keeping only three tabs open (and although it still struggles with some of the largest spreadsheets and dashboards at work, it is a delightful second-screen device).
But I think the M1 still has an edge because of the vertical integration between software and hardware -- if you previously wrote an app against Core Video or Metal, it should get hardware acceleration once you recompile it for the M1. Whereas if similar accelerators were exposed in Linux or Windows you'd probably have to write against a new custom API/library (such as nvenc).
There is nothing bad that someone stayed behind. Someone has to stay behind if we want progress. No reason to insult predecessors.
Planc and Dirac weren't making fun of Newton. Minitel was used in France as some simple version of Internet years before Internet was created. Was Minitel bad? Should we make fun of it?
Was Minitel standing in its underwear in front of Internet? Has Dirac stolen Newton pants?
Intel made a lot of great stuff. Now Apple is making great stuff (or, to be precise, TSMC does). Next year someone else is going to make great stuff. Maybe Apple maybe not.
Intel and AMD chips are not that far behind, depending on what benchmark is used, even though they use older 14/10 or 7nm technology.
I also really hate the concept of not being able to upgrade RAM - If I use Mac and I need more RAM I need to throw away my old Apple laptop and buy a new one. So I have to spend way more money, not to mention impact on environment.
So, today, I am still choosing Intel/AMD computer, I can install whatever operating system I wish and upgrade my system to whatever configuration I dream about and I can afford now or in future.
A ram upgrade will never compensate for a less efficient memory architecture.
There is no CPU architecture other than Apple’s that will make your laptop faster because AMD and Intel are thermally limited, so there is nothing to upgrade to.
The only situation in which you are right, is if you are talking about a high end desktop tower.
If my program needs 64GB of ram, your 8GB highly efficient "memory architecture" can do nothing about it...you have to know that even my gpu has 16GB alone. But i have to say that 8GB are probably enough for your Chrome Browser.
Agree largely with the rest of your comment, but this particular event was more than 23 years ago. I'm sure a good number of readers of this thread is younger than that!
We aren't running ARM on our desktops because there aren't ARM desktops for us to use. The market jumps from low-end boards like the Pi and Pine's lineups - which are excellent for what they are, but suffer from poor performance and I/O - to the high end where ex. https://www.asacomputers.com/Cavium-ThunderX-ARM.html looks like a nice system but $2,820 is a bit of a barrier to entry.
Indeed, while ARM based Linux desktops hit the price and energy efficiency on the nose, they aren't going to wow anyone with performance.
Be careful what you wish for. X86 is the most open architecture right now. ARM, on the other hand, has locked bootloaders, messy device trees and so on. You can see that virtually every Linux distro running on an ARM SOC needs some specific tweaks.
The reason PCs are still open is historical, they started like this and people can't really see PCs any other way but they're a dying breed and whatever comes after is a different game with different rules. We'll have some niche manufacturer building something more open for the small segment of enthusiasts.
The fact that so many people casually say how a platform is open or flexible because if you jump through enough hoops you get to make some minor modifications to your device, or applauding the openness of a system for allowing you to sideload standard apps is the reason manufacturers don't see a need to propagate the PC philosophy anymore. Too many users grew up with this new model so they don't see the loss.
None of the devices you mention are user serviceable in any way hardware-wise, the software is almost always locked onto the device and there's no easy, or even supported way of changing it. And if you do change it you just get a small variation of the same software. Your phone may get a slightly degooglified version of Android but it's unlikely you'll get it to run Ubuntu, Windows, or a FreeBSD without a massive investment in time and knowledge.
I don't expect or even really want to swap hardware on these devices: if that was a priority for me I would get a PC, Arduino, Raspberry Pi, etc. I get the desire to tinker and customize, but from where I'm sitting I have plenty of opportunity to do that.
I also don't expect the device manufacturers to do the work to make Ubuntu or FreeBSD run. They didn't do that for PCs, the community did.
> We'll have some niche manufacturer building something more open for the small segment of enthusiasts.
I would argue that's exactly what Raspberry Pi and Arduino is.
This is why you need rpi specific distro builds.
You might want to take a look at PowerPC/OpenPOWER.
ARM not being very open is mostly because there's been limited development interest in it, aside from specialised devices, and those companies have no real need or interest to have open things. Doesn't mean ARM itself can't evolve past it.
That's an extremely weird thing to say...
x86 is not "Open" in any fashion. You can't license x86 even for a fee, let alone free.
While ARM isn't open either, it is by comparison far more available to license at reasonable cost.
> ARM, on the other hand, has locked bootloaders
It also has open/ unlocked boot loaders. One of the advantages of licensable technology like this is you can have a lot of different architectures with the same underlying core. With x86, you get whatever the Intel/ AMD duopoly decide to ship you.
OP is not talking about being able to manufacture x86 CPUs kind of open, that's useless for most folks.
They're talking about being able to install any OS, drivers being fairly universal, no device specific out of tree kernel patches for the vast majority of PCs, no device trees etc.
I mentioned that as well though. Using your goofy definition of Open, x86 is no more "Open" than ARM. Pinebook and Raspberry Pi both use no bullshit ARM setups. With ARM, how accessible the architecture is is 100% dependent on the implementation which vary greatly by manufacturer.
If I buy any x86 PC, chances are I can install mainline Linux on it. Picking up any particular recent ARM SoC and the chance of doing so is much, much less. The M1 included, after all.
> Pinebook and Raspberry Pi both use no bullshit ARM setups.
Both of these took considerably longer to get mainline support that your average x86 hardware needs.
Not sure what you are thinking here. It's taken 25 years so far and we still have tons of x86 hardware which isn't supported well in Linux. There is a reason Dell ships a specific laptop for developers with Linux pre-installed.
And yet, even Windows DELL laptops are more likely to run mainline Linux than any average ARM SoC, even including the ones targeted at developers.
I am no fan of x86, but trading what we have now for an M1 is a net loss in terms of our ability to install anything we want as users, no question.
> And yet, even Windows DELL laptops are more likely to run mainline Linux than any average ARM SoC, even including the ones targeted at developers.
This is more about inertia and developer resources than openness. The more developers you have on ARM, the faster device and software support will come.
> I am no fan of x86, but trading what we have now for an M1 is a net loss in terms of our ability to install anything we want as users, no question.
You've changed the question here. We were speaking about ARM in general, not the M1.
The second is if there's linux port for that hardware, which it isn't. There's a project underway, seems likely that the CPU (which is still AARCH64) will work, and likely networking. However to make a usable desktop you'll need a GPU which is much harder.
They're ARM SoCs, meaning that you can't just download the Ubuntu ARM ISO[1] and run it on either device like you could with an x86 ISO and an x86 machine.
To run Ubuntu or any Linux distribution, you need to use images that were specially crafted for either SoC.
This is because ARM SoCs use custom bootloaders and don't have hardware on enumerable buses. Each SoC model is essentially a unique hardware configuration that needs an individual Linux port.
ARM servers are more similar to x86 machines in that they use UEFI and have enumerable buses.
As long as machines are shipped that use ARM SoCs, Linux support for all of them will be quite difficult to achieve[2] compared to x86 machines.
[1] https://ubuntu.com/download/server/arm
[2] https://elinux.org/images/a/ad/Arm-soc-checklist.pdf
(I also think it's a bit too early to say definitively that no ARM-based Apple computer will ever be able to boot into Linux. I don't think Apple has any motivation to make that easy, but that may not be the same as making it impossible.)
> I also think it's a bit too early to say definitively that no ARM-based Apple computer will ever be able to boot into Linux.
There's a big difference between a SoC running a Linux fork and actually running Linux distributions in the same sense that we can on x86 computers or ARM servers.
[1] https://docs.microsoft.com/en-us/windows-hardware/drivers/br...
[1] https://www.xda-developers.com/microsofts-debug-mode-flaw-an...
Seriously? You need a blob from Broadcom to even boot the thing, and you call that a no bullshit ARM setup? I am astounded.
Are the FLOSS drivers for Nvidia's cards great now? How many years have people been hacking on that project to make them work as well as they do?
What people are describing as "Open" is actually an enormous amount of work on the part of developers and the grudging support of hardware vendors after decades of struggle.
Very few actually.
> Are the FLOSS drivers for Nvidia's cards great now?
Don't shift goalposts, we are talking about CPU support, not GPUs... GPU support on ARM is anyway not that great to begin with.
> What people are describing as "Open" is actually an enormous amount of work on the part of developers and the grudging support of hardware vendors after decades of struggle.
Intel has Linux dedicated teams at least, pushing FOSS code every single day. Most ARM licensees could not care one bit about FOSS.
Linux already supports ARM CPUs and has for some time. The challenge is drivers. Getting video/ wifi/ etc running is the part that makes devices actually useful. The Broadcom Blob the poster I replied to was referring to is not for supporting ARM. Very similar to the Nvidia blob many people install (often unknowingly) when they install Linux on a system with Nvidia.
> Intel has Linux dedicated teams at least, pushing FOSS code every single day. Most ARM licensees could not care one bit about FOSS.
Intel didn't give a shit about Linux for how many years? Companies don't care about FOSS until it has potential to affect their bottom line. It was only after Linux servers were running half the Internet, that Intel stepped up and started supporting Linux.
I'm guessing you missed the first couple decades of Linux adoption. Linux on ARM is just a fast motion repeat of the early days of Linux on x86.
RISC-V: Not really. The closest thing is probably https://www.sifive.com/boards/hifive-unmatched
at prohibitive costs. So not really an option...
Can you please elaborate?
> exceptions to the above include i2c (unusual, and taken care of by i2c-sensors, which uses good heuristics to "probe" devices from userspace) and the ISA bus and its derivatives such as Compact Flash and IDE. even PCMCIA got sufficient advances to auto-identify devices from userspace at runtime.
> so as a general rule, supporting a new x86-based piece of hardware is a piece of piss. get datasheet or reverse-engineer, drop it in, it's got BIOS, ACPI, USB, PCIe, SATA, wow big deal, job done. also as a general rule, hardware that conforms to x86-motherboard-like layouts such as the various powerpc architectures are along the same lines.
> so here, device tree is a real easy thing to add, and to some extent a "nice-to-have". i.e. it's not really essential to have device tree on top of something where 99% of the peripherals can describe themselves dynamically over their bus architecture when they're plugged in!
> now let's look at the ARM world.
> * is there a BIOS? no. so all the boot-up procedures including ultra-low-level stuff like DDR3 RAM timings initialisation, which is normally the job of the BIOS - must be taken care of BY YOU (usually in u-boot) and it must be done SPECIFICALLY CUSTOMISED EACH AND EVERY SINGLE TIME FOR EVERY SINGLE SPECIFIC HARDWARE COMBINATION.
> * is there ACPI present? no. so anything related to power management, fans (if there are any), temperature detection (if there is any), all of that must be taken care of BY YOU.
> * what about the devices? here's where it becomes absolute hell on earth as far as attempting to "streamline" the linux kernel into a "one size fits all" monolithic package.
> the classic example i give here is the HTC Universal, which was a device that, after 3 years of dedicated reverse-engineering, finally had fully-working hardware with the exception of write to its on-board NAND. the reason for the complexity is in the hardware design, where not even 110 GPIO pins of the PXA270 were enough to cover all of the peripherals, so they had to use a custom ASIC with an additional 64 GPIO pins. it turned out that that wasn't enough either, so in desperation the designers used the 16 GPIO pins of the Ericsson 3G Radio ROM, in order to do basic things like switch on the camera flash LED.
> the point is: each device that's designed using an ARM processor is COMPLETELY AND UTTERLY DIFFERENT from any other device in the world.
https://lists.debian.org/debian-arm/2013/05/msg00009.html
This post is so outdated, that ARM servers' companies had the time to have everything, then still die (most of them). And then revive with the AWS Graviton2 and other Neoverse chips.
"Fastest chip" uhh well no..
"Fast for this one thing", yeah it's 0.1s faster opening that program no one ever said was slow.
"It has good battery life"
So this is the new narrative? Are nerds constantly fighting back against Apple marketing and they are adapting?
Usually their marketing team just insists over and over. Maybe because this is actually measurable (unlike privacy and security) their legal team is holding the marketing team back.
I don't think that is an indictment of Linux. It is likely if the specs for the M1 were released, Linux could wring more performance out of the M1 than MacOS. At least in some workloads.
I think it’s perfectly legitimate to raise concerns around Apple software because it’s one area where Apple has faced sustained criticism from its users and has done very little to remedy the situation. So if the software that runs on top of the M1 is shakey, it’s entirely realistic to expect these “warts” to be with the platform for a while.
If there are software issues, address them as such.
Why are we giving Apple a pass on RAM being stuck at 8GB and SSD being stuck at 128GB (only very recently bumping to 256) for almost a decade. Not to mention some of us get stuck on 8GB 128GB machines because our workplaces only buy the base "Pro" machines, refusing the drop the extra few hundred to make it what Gruber considers the base level.
SSDs are no longer expensive, go on Amazon and see how much 1TB M2s cost these days.
Work provides i7/i9 with maxed out RAM and honestly don't see a night and day difference in performance.
Sure, if you can bear living in a prison complex that has shitty window management and a petty tyrant of a warden.
Maybe I'm wrong though - it might depend on the thread. Whenever I put out an anti-Apple sentiment on a pro-Apple article thread I usually get massive amounts of disagreement.
My own dislike for Apple goes way beyond their protectionist policies. Maybe if I didn't hate the Mac UI with that hideous global menu bar and horrid window management, I could be more forgiving towards their politics. Heck, I used Windows as my main OS for decades until I finally switched to Manjaro and XFCE for the sheer convenience of running Docker and npm/yarn smoothly without all the Windows cruft slowing me down all the time.
It's funny though - I do less tinkering on my Linux desktop than I ever did on Windows (or my Mac), since everything just works the way I want it to in Manjaro. I think my first week using it, I had to figure out how to write an xdotools script to get window snapping in XFCE. Never had to touch it since. My Mac has broken things like Karabiner, Better Touch Tool, Little Snitch and so forth after many updates - so much so that I just stopped reinstall them. Luckily, I only have to use it once in a while for doing the occasional Mac/iOS thing.
I want access to my boot volume. I don't want cloud anything.
So Apple's technical prowess now makes me sad. I can either use an OS that doesn't head further every year in a direction I hate, or I can have a fast computer.
Market opportunity here for a peripheral that's just an adhesive strip you put above your touch bar, and it outputs heat corresponding to CPU usage ;)
"Safari is faster than Chrome" NOPE Now that IE is gone, Safari is the new IE by being the last not-ever-green browser always behind in features. So many benchmarks again and again proved Chrome and Firefox to be faster than Safari.
"iOS is faster than Android" NOPE It just depends on the device. Of course the latest iPhone is faster than a 300$ mid-low range Android device. Pick the top Android devices vs iPhones and you will see it's a tough race with no clear winner.
In the benchmarks I've seen (https://arstechnica.com/gadgets/2020/11/google-chrome-is-ava...) and the ones I've run myself, Safari annihilates Chrome.
> Pick the top Android devices vs iPhones and you will see it's a tough race with no clear winner.
No, all Android phones are clearly slower than the iPhone. https://www.tomsguide.com/news/iphone-12-benchmarks-this-des...
2. The new chip does lots of little things waaaay faster than the old chip does a handful of big things.
3. Apple figured out how to do a lot more things before they need to. Like if you did all of the math problems in your textbook ahead of time and then when the teacher eventually gives you the assignment you hand in the ones they asked for.
I found the following article enlightening:
https://screenrant.com/apple-silicon-m1-mac-macbook-chip-soc...
Edit: ignore that previous link, way better is the one mentioned by another poster:
https://debugger.medium.com/why-is-apples-m1-chip-so-fast-32...