Is Silicon M1 a Game Change?

[thejadedmonkey]

Could you expand a bit more on this? There are some differences, sure, but I have to say that this statement surprises me a bit. On a conceptual level, I don't see much of a contrast between the two. I wouldn't approach writing code for an ARM platform any different than writing code for an x86 platform (which is incidentally the main reason why existing software runs great on both x86 and ARM).

It's not different. I wrote code for x68 for years, and for the past few years have been writing exclusively for ARM (phones). The only difference is maybe compilers, but the principles of software engineering (readability, how to make network calls, etc) remain the same.

 

[leman]

It's not different. I wrote code for x68 for years, and for the past few years have been writing exclusively for ARM (phones). The only difference is maybe compilers, but the principles of software engineering (readability, how to make network calls, etc) remain the same.

That's also my sentiment, which is I was curious to see what @gustavopi means.

 

[UBS28]

What a load of nonsense. Mac and iOS always used the same OS (OS X) and they've shared over 90% of their source code from the very beginning (the only difference being the UI subsystem). And yeah, people have been claiming these nonsensical things for a while, and 10 years later there is still no sign of macOS and iOS merging.

Have you not noticed all those iOS features like notifications?

 

[leman]

Have you not noticed all those iOS features like notifications?

How is that an iOS feature? I was using Growl notifications on a Mac since Leopard, years before iOS has been released. Growl was so popular and useful that all polished apps had notification support. In fact, Apples Notification Center looks and works a lot like Growl, so they probably got inspired by the project.

And of course there are things in macOS that are similar to iOS. Take the new control center that is copied straight from iOS. But I don’t see how this is „turning macOS into iOS“. It’s simply taking a useful feature and adapting into a new platform. Frankly, I approve of all Apple platforms sharing visual language where it is possible, because it only makes sense. But just because two things look similar it doesn’t mean that they are the same or have the same functionality. In fact, I think it makes more sense to argue that iOS is becoming more like macOS and not the other way around - iOS has gained a limited form of multitasking, file management, smb file sharing, Keyboard/mouse support while macOS has lost… nothing.

 

[trevpimp]

It's a game changer for everything especially when they put the M1 in the 12.9 iPad Pro. Apple would take compatibility beyond software and hardware with the M1 being put in both iPad and Macs. I feel like they have something up their sleeve with incoming head turning products.

As far as computing I feel like they are way ahead of the game in which they have the first stand in the computer market. I know Dell, Microsoft, Toshiba etc will have to follow up with Apple choosing whats available before other companies.

 

[gustavopi]

Could you expand a bit more on this? There are some differences, sure, but I have to say that this statement surprises me a bit. On a conceptual level, I don't see much of a contrast between the two. I wouldn't approach writing code for an ARM platform any different than writing code for an x86 platform (which is incidentally the main reason why existing software runs great on both x86 and ARM).

I was afraid you ask me this... ok, I will try to explain with a basic diference between the architectures. x86 is CISC, therefore is focused in to have complex instructions inside the chip (not all have but we are considering the x86 family) so you will have less memory usage (RAM and the executable size) and will be easier to compile the code, and to write low level code (assembly). The RISC is focused in optimized project - less transistors, small size, less power consumption, it won't have those instructions, but will have another features and the freedom for you to make a better solutions, is a more flexible architecture. Of course most of us won't write in assembly or develop compilers, but if you are game coder, you will be aware to the fact that might have new different solutions available for rendering, calculating or whatever you need to include in your project. Things that might increase file size and memory usage are not a critical problem these days in comparison with the 90's, we wish a fast a stable game full of resources. So you might, as high level programmer, change your focus to this aspects when developing.

I think the will take a while to change how we make programs so we can get the most of M1. One example is crypto mining: as long these new Macs are powerful, they are still losing for traditional x86 PC's (I don't know but I bet the apps are running under Roseta2). I bet we will have Macs (or some M1/M2 based hardware) in the crypto market in the years to come.

 

[JMacHack]

It’s a canary in a coal mine at least. The M1, first low-end, low-power Mac chip that Apple released, is trading blows with mid to high end x86 processors that use at least double the power, and has the best integrated graphics bar none. Apple made a processor that people did not believe could be this performant for its efficiency.

After it was announced, Nvidia and AMD announced plans to make ARM cpus, Samsung recently talked about their own cpus, and I’m sure there are more companies that are going to try their hand at ARM cpu making.

For the moment, x86 is the king of raw desktop performance. But I believe that it won’t be for much longer.

 

[gustavopi]

Just some curious fact about game change involving RISC vs CISC processors:

Back in the early 2000's I bought my iMac green transparent in a promotion, (new iMacs were coming). I was working in a company that was just bought PC's with Pentium II. My working PC was a customized one with SCSI HD and a Voodoo GPU over a powerful motherboard - a hell of a PC!. As my iMac was a "portable" desktop (I could easily carry), I taken it to my job to show to my PCzists friends and perform a comparison side by side... why not. The claim of the advertising was twice speed... average. Well, with 333MHz it was really close to the turbo V8 PC with 400MHz. But the real test was with a brand new Pentium 3 550MHz, a common PC desktop as my iMac. The main test was rotating a picture in Photoshop, and the Mac did in half time! Not to mention that beije desktop full of cables was a dinosaur side by the iMac (I miss that iMac...). RISC beat badly the CISC at that time, but Intel never needed to win drag races to win the market, right? Something to think about...

The direct competition of PPC G3, the leader, the Intel Pentium II: big, fat, hot as hell and slow (at list with P1 bugs fixed)

The answer from Intel to PPC Macs: we are getting better, even 2x behind yet...

The IBM awesome, PowerPC 7xx family: quickly, thin, cold, the same of the laptops...

Could you imagine, before 1997, a laptop with the same CPU of a desktop??? That was one game change. I can admit M1 can be too, but not in the same level.

 

[machinesworking]

There are plenty of signs for anything if you look for them. That's how conspiracy theories are born.

OK this again is what I'm chiming in about, this is a false equivalency, thinking Apple might be getting multi touch for Mac OS is not a conspiracy theory, it's not a comparable line of thinking. You're jumping to major denial of any other possibility than the one you offer for basically speculation on all of our parts, none of us actually know Apples road map, and if we did we would be under serious NDA's, we would not be here speculating at all. You're invested in your opinion to the point to where you're accusing people of thinking like conspiracy theorists and their ideas are loads of nonsense etc. You're coming back on speculation with certainty, which is equally unrealistic, since you can't know, and if you did, you wouldn't be here.

 

[quarkysg]

It's not different. I wrote code for x68 for years, and for the past few years have been writing exclusively for ARM (phones). The only difference is maybe compilers, but the principles of software engineering (readability, how to make network calls, etc) remain the same.

Unless the code are kernel/OS level code, I don’t think there’re much differences between ISA for app layer codes. These are all abstracted away by the language and OS APIs, and the compilers takes care code generations for the ISA.

Nowadays developers target platforms instead of ISAs isn’t it?

From my limited experience with Linux drivers, most of the time, we just need to take care of endianess. Otherwise it’s just C codes calling kernel APIs. I don’t really bother whether it’s to target ARM or x86, as the APIs takes care of it mostly.

 

[leman]

I was afraid you ask me this... ok, I will try to explain with a basic diference between the architectures. x86 is CISC, therefore is focused in to have complex instructions inside the chip (not all have but we are considering the x86 family) so you will have less memory usage (RAM and the executable size) and will be easier to compile the code, and to write low level code (assembly). The RISC is focused in optimized project - less transistors, small size, less power consumption, it won't have those instructions, but will have another features and the freedom for you to make a better solutions, is a more flexible architecture.

Yeah, I know about RISC and CISC. I frequently see explanations like the one you give here but in the end, if you look at all this in more detail, I don’t think there is much left of this argument. The main difference between Aarch64 and x86 is that the first is a load-store architecture with fixed instruction length and the second is a register memory architecture with variable instruction length. This means that you can occasionally encode a sequence of operations in a slightly more compact way in x86, and that’s about it… and even then it mostly applies to toy examples. When you look at real world code, especially optimized one, there is no significant difference between x86 and Aarch64 binary sizes.

As to complex vs. simple instructions… frankly, I don’t even know where do these misconceptions came from. ARM has instructions that store/load multiple registers at once, it has matrix multiplication, auto-increment addressing modes, pointer encryption, half-float operations, interleaved simd loads and stores, fully featured horizontal vector operations… in fact, I can’t think of any instruction that x86 has that ARM lacks (except maybe the weird old-school CISC instructions which are dead slow and never used in modern code).

In the end, Aarch64 is a modern instruction set that has been meticulously designed to simplify the lives of both the CPU designer and the compiler writer as well as enable high-performance out of order execution. And x86 is a very old design that has been continuously adding more modern features on top of its legacy core.

Of course most of us won't write in assembly or develop compilers, but if you are game coder, you will be aware to the fact that might have new different solutions available for rendering, calculating or whatever you need to include in your project. Things that might increase file size and memory usage are not a critical problem these days in comparison with the 90's, we wish a fast a stable game full of resources. So you might, as high level programmer, change your focus to this aspects when developing.

I think I understand what you are trying to say here. Aarch64 includes some new technologies, and it obviously takes skill and experience to make good use of them. If you are an experienced SIMD programmer on an x86 platform, it will get you some time to get good at utilizing ARM Neon and ARM SVE SIMD. But that is not x86 vs ARM, that's just learning the platform. On the conceptual level, there is no difference.

Apple Silicon of course brings other features. Its focus on heterogenous programming for example enables one to combine the processing powers of multiple devices (CPU, NPU, GPU) in a way that is not really possible with a traditional pooled memory model. So there is also a certain learning curve here while we figure out what is possible.

 

[leman]

OK this again is what I'm chiming in about, this is a false equivalency, thinking Apple might be getting multi touch for Mac OS is not a conspiracy theory, it's not a comparable line of thinking. You're jumping to major denial of any other possibility than the one you offer for basically speculation on all of our parts, none of us actually know Apples road map, and if we did we would be under serious NDA's, we would not be here speculating at all. You're invested in your opinion to the point to where you're accusing people of thinking like conspiracy theorists and their ideas are loads of nonsense etc. You're coming back on speculation with certainty, which is equally unrealistic, since you can't know, and if you did, you wouldn't be here.

Sorry if I gave you that impression. My point is simple: people have been prediction that macOS will turn into iOS for the last 10 yeas - ever since Mac App Store launched on Mountain Lion. Every single time a new OS X version was released, there was an outcry of prediction that this was the "last real OS X" and soon all Macs will be essentially big iPads with limited functionality. But in the end, all we got was some cosmetic unification and API harmonization, that's it. Mac is still a Mac and an iPhone is still an iPhone.

It is true, I think that the chance of Apple planning touch support to macOS is very slim. But it's not because I am in denial agains any other opinion. It's because Apple said themselves — on multiple occasions — that they plan no such thing, because macOS programming frameworks are not set up with touch inputs in mind and there are no signs that they are being overhauled to offer such support. Apple does some things long-term. If you know where to look, their transition to ARM for example was clear years ago (you just need to look at subtle changes in ABI, Metal, etc.). The claims about iOSification of macOS on the other hand I don't take seriously — because these claims are based on superfluous, cosmetic things and tend to completely ignore the underlying implementation.

 

[Realityck]

I voted almost, because while the M1 is quite respectable, its not competitive with better intel based mac configurations.

 

[altaic]

I’m waiting to vote after the WWDC keynote.

 

[machinesworking]

Sorry if I gave you that impression. My point is simple: people have been prediction that macOS will turn into iOS for the last 10 yeas - ever since Mac App Store launched on Mountain Lion. Every single time a new OS X version was released, there was an outcry of prediction that this was the "last real OS X" and soon all Macs will be essentially big iPads with limited functionality. But in the end, all we got was some cosmetic unification and API harmonization, that's it. Mac is still a Mac and an iPhone is still an iPhone.

It is true, I think that the chance of Apple planning touch support to macOS is very slim. But it's not because I am in denial agains any other opinion. It's because Apple said themselves — on multiple occasions — that they plan no such thing, because macOS programming frameworks are not set up with touch inputs in mind and there are no signs that they are being overhauled to offer such support. Apple does some things long-term. If you know where to look, their transition to ARM for example was clear years ago (you just need to look at subtle changes in ABI, Metal, etc.). The claims about iOSification of macOS on the other hand I don't take seriously — because these claims are based on superfluous, cosmetic things and tend to completely ignore the underlying implementation.

Sure, but I think there's a qualitative difference in the reasons one might be saying it. For years people have been saying Mac OS was becoming iOS in a "The Sky is Falling!" way. The OS was becoming a "toy" OS etc. Nothing really big to say that was true ever happened. I'm partial to multi touch interfaces so take that for what it is, I almost bought an Asus Zenbook last November until the M1 was shown. That said, iOS apps now run on Mac OS, Mac OS now runs on Apple Silicon, the same chips in the iPad Pro are in the iMac. All those things weren't true a few months ago, I 100% get why someone would think it's likely that Mac OS gets multi touch support.
On the other hand the new rumor is Apple own pro products are being ported to iOS, namely Logic and FCP.

Some of this all seems a little too like the situation with the one button apple mouse. Apple disregarded complex mice until they didn't

 

[mi7chy]

Lacking current AAA games so not game change.

 

[xraydoc]

I voted for game changer, but that doesn't necessarily mean it's all good.

While certainly the M1 and successors have industry-leading power-per-watt ratings, and the new unified architecture leads to very low latencies which helps make things feel even faster, it also likely means a return to the completely closed architecture of the Mac -- a sealed box, computer-as-appliance paradigm that was championed by Steve Jobs (even more so than we have today) without upgradable/replaceable parts.

I'm fully anticipating the next Mac Pro to be a G4 Cube/Trashcan Mac Pro-type machine with little user-accessible internal space. Maybe a couple of SSD slots for internal storage expansion, but not much else.

 

[Blue Quark]

“Always in motion, the future is.”
— Yoda

I think Apple felt the need to justify some of that investment they've been making in R&D. I'm also a fan of competition, so hopefully this spurs the Intel / AMD juggernaut to improve things. There's also a considerable degree of threat coming from exploits against (mostly Intel's) speculative execution functionality, so maybe Apple figured they could take the lessons learned from, let's call it "what needs to come out the other end of the pipe" and find better ways of making that all happen.

 

[dmccloud]

It is not a game changer. We have had the “M1” in the iPad Pro’s for years.

It was predicted a long time ago that Mac and iOS devices would merge at some point. First OS X got iOS like and now they are using the chips from iOS devices.

Not going to happen - during WWDC last year and even in some interviews leading up to the actual release of the M1-based Macs, Apple repeatedly stated that they would continue to keep iOS and Mac OS as separate platforms. If anything, moving the iPad Pro to the M1 was more about distinguishing the Pro from the rest of the iPad lineup than it was about any sort of merger between the two operating systems.

 

[jdb8167]

I'm fully anticipating the next Mac Pro to be a G4 Cube/Trashcan Mac Pro-type machine with little user-accessible internal space. Maybe a couple of SSD slots for internal storage expansion, but not much else.

That won’t fly any better than the 2013 Mac Pro. I love my 2013 Mac Pro but I’m in the minority on that. A Mac Pro without expansion won’t sell.

 

[xraydoc]

That won’t fly any better than the 2013 Mac Pro. I love my 2013 Mac Pro but I’m in the minority on that. A Mac Pro without expansion won’t sell.

I agree. However, I'm not sure if Apple sees it that way.
In the current Mac Pro, what's the expansion for? GPU cards, drives and RAM.
I'm expecting an absence of GPUs, since it'll be integrated into the CPU as it is on the M1. Same with RAM. So we're left with maybe a couple SSD slots for additional storage (main SSD will be integrated like on the M1 SOC).

 

[4sallypat]

It’s a canary in a coal mine at least. The M1, first low-end, low-power Mac chip that Apple released, is trading blows with mid to high end x86 processors that use at least double the power, and has the best integrated graphics bar none. Apple made a processor that people did not believe could be this performant for its efficiency.

After it was announced, Nvidia and AMD announced plans to make ARM cpus, Samsung recently talked about their own cpus, and I’m sure there are more companies that are going to try their hand at ARM cpu making.

For the moment, x86 is the king of raw desktop performance. But I believe that it won’t be for much longer.

Yes, agreed, the M1 was a test and Apple passed with flying colors.

So much so that my measly 8GB Unified Memory (base) M1 Mini and M1 iMac are blowing away 15 years worth of Intel Macs with double, triple, and quadruple DDR RAM!

For a "low end, low power" Apple Silicon, I am simply amazed at how it works in my daily life!

Those that keep talking about Pro versions - 32GB of Unified Memory - M2x, M1x and P1, etc - hate to say it but 90% of us are normal users will find the M1 to be more than sufficient.

 

[BigMcGuire]

Been a game changer for me. For the first time ever I installed Windows 10 Pro on my lap, set up VS 2019 Pro with a massive project and compiled it - without hearing a single fan. And ... on my lap, did I mention on my lap? Even with a lap desk, my work provided 2019 MBP 15' was overheating me and screaming fans (having to be plugged in).

The UI experience alone is shocking - instant actions, smooth operations ... love it.

While I look forward to 32+GB M1X or M2s - this current M1 does everything I need and more - all the games I want to play - no problems. I've got Age of Empires HD working on CrossOver and Starcraft II - to a 4k external monitor. It's almost as good as my RX 580 eGPU without the stutter!

 

[jdb8167]

I agree. However, I'm not sure if Apple sees it that way.
In the current Mac Pro, what's the expansion for? GPU cards, drives and RAM.
I'm expecting an absence of GPUs, since it'll be integrated into the CPU as it is on the M1. Same with RAM. So we're left with maybe a couple SSD slots for additional storage (main SSD will be integrated like on the M1 SOC).

I don’t think so. There is no reason to disallow RAM upgrades. The current Mac Pro has tremendous memory bandwidth. 140 GB/s according to Apple. That is more than twice as fast as the M1.

The SSDs on the current Mac Pro are faster than the SSDs on the M1 Macs. So again there is no reason to change that design.

As for slots, Apple is rumored to be creating their own GPU. There is no reason why they couldn’t put that on an MPX card and allow users to upgrade to newer versions over time. I also am not sure that Apple is going to block AMD graphics cards on the Mac Pro. Having access to very high end GPUs is something that Apple might not be able to do in house.

 

[carlvsam]

Just some curious fact about game change involving RISC vs CISC processors:

Back in the early 2000's I bought my iMac green transparent in a promotion, (new iMacs were coming). I was working in a company that was just bought PC's with Pentium II. My working PC was a customized one with SCSI HD and a Voodoo GPU over a powerful motherboard - a hell of a PC!. As my iMac was a "portable" desktop (I could easily carry), I taken it to my job to show to my PCzists friends and perform a comparison side by side... why not. The claim of the advertising was twice speed... average. Well, with 333MHz it was really close to the turbo V8 PC with 400MHz. But the real test was with a brand new Pentium 3 550MHz, a common PC desktop as my iMac. The main test was rotating a picture in Photoshop, and the Mac did in half time! Not to mention that beije desktop full of cables was a dinosaur side by the iMac (I miss that iMac...). RISC beat badly the CISC at that time, but Intel never needed to win drag races to win the market, right? Something to think about...

The direct competition of PPC G3, the leader, the Intel Pentium II: big, fat, hot as hell and slow (at list with P1 bugs fixed)
View attachment 1786570

The answer from Intel to PPC Macs: we are getting better, even 2x behind yet...

The IBM awesome, PowerPC 7xx family: quickly, thin, cold, the same of the laptops...
View attachment 1786572
Could you imagine, before 1997, a laptop with the same CPU of a desktop??? That was one game change. I can admit M1 can be too, but not in the same level.

If I remember it correctly, those rectangular Intel CPU modules were actually an attempt to have a proprietary slot standard to prevent the use of AMD Athlon CPUs - which were compatible with the "square" sockets Intel had been using (this was around the first time AMD had produced CPUs that could compete with what Intel was cranking out.) Fortunately, what Intel try to do failed, and they had to go back to the more conventional sockets.

As for the PPC CPUs, I believe one of the main reasons Apple stopped using them was the same reasons they decided to replace Intel CPUs: as the CPUs got more powerful, the thermals became more of a problem (for instance, the fasted G5 PowerPC ran so hot that the CPU had to be liquid-cooled.)

Ironically, at the same time Apple is abandoning Intel CPUs, AMD has mostly surpassed them in X86 performance again with their Ryzen CPUs …?