Devious marketing of M2

[Gudi]

Humor us then, please. When you write "RISC" and "CISC", what is that you mean exactly?

I told you multiple times. It's two fundamentally different kinds to build processors. One way more energy efficient than the other. And since you can't build smartphones, the predominant kind of computers, with CISC chips, x86 is an obsolete technology and all companies dependent on its continued existence are doomed to go bankrupt. Sorry if the future is not to your liking.

 

[Xiao_Xi]

It's two fundamentally different kinds to build processors. One way more energy efficient than the other. And since you can't build smartphones, the predominant kind of computers, with CISC chips, x86 is an obsolete technology and all companies dependent on its continued existence are doomed to go bankrupt.

Why did Apple switch from RISV-based PowerPC to CISC-based x86?

Apple Makes the Switch: iMac G5 vs. iMac Core Duo

www.anandtech.com

 

[leman]

I told you multiple times. It's two fundamentally different kinds to build processors. One way more energy efficient than the other. And since you can't build smartphones, the predominant kind of computers, with CISC chips, x86 is an obsolete technology and all companies dependent on its continued existence are doomed to go bankrupt. Sorry if the future is not to your liking.

Ok, so RISC "good" and CISC "bad". Clare to clarify what these "kinds to build processors" are and how exactly are they "fundamentally different"? Because so far you have failed to produce anything remotely of substance except general platitudes. You can accuse me of being clueless all you want, but at least I have provided a historical account for these terms and explained why they don't really matter in modern computing. Let's have a constructive technical discussion. You could start from explaining the specific properties of RISC which allow it to have all these great advantages you claim it to have.

 

[Joe Dohn]

Not quite sure what you mean by the "didn't call it Rosetta out of nowhere"? Rosetta is a reference to the Rosetta Stone, which is an artefact that has been used to decipher Egyptian script. You can see it in London. It's very impressive.

You don't need to lecture me on the original Rosetta stone. I know what it is.
What I meant, obviously, is that they named THIS Rosetta 2 after the original Rosetta to further emphasize that this translation layer is temporary, like the original one was. No one anywhere (including on this forum) is expecting it to become a permanent part of the system, or to remain for 15+ years or any long-term timeframe.

 

[jdb8167]

Ok, so RISC "good" and CISC "bad". Clare to clarify what these "kinds to build processors" are and how exactly are they "fundamentally different"? Because so far you have failed to produce anything remotely of substance except general platitudes. You can accuse me of being clueless all you want, but at least I have provided a historical account for these terms and explained why they don't really matter in modern computing. Let's have a constructive technical discussion. You could start from explaining the specific properties of RISC which allow it to have all these great advantages you claim it to have.

He could also answer the question, is a superscalar processor still RISC? What is micro-op fusion. If a CPU uses micro-op fusion, is it still RISC? I'll wait.

 

[leman]

What I meant, obviously, is that they named THIS Rosetta 2 after the original Rosetta to further emphasize that this translation layer is temporary, like the original one was.

That's a fairly far fetched conclusion.

No one anywhere (including on this forum) is expecting it to become a permanent part of the system, or to remain for 15+ years or any long-term timeframe.

Nothing is permanent. Certainly not with Apple. They deprecate technologies all the time. Like they deprecated Carbon, 32bit and many of their frameworks. Right now they are preparing to deprecate AppKit and UIKit in favour of SwiftUI, and they will probably drop ObjectiveC as well in a fairly near future. Still, does it mean that macOS has no useable software?

It's very much possible that in five years none of the current Intel-based games will run on a newest Macs. Similarly, none of the current ARM software will probably run on a Mac in 15 years (if there are still Macs). These are hardly the reasons to pretend that software doesn't exist or is unusable.

 

[leman]

He could also answer the question, is a superscalar processor still RISC? What is micro-op fusion. If a CPU uses micro-op fusion, is it still RISC? I'll wait.

We can also start with basics, like "how many registers does a modern CPU have"? Already that is a fun one.

 

[jdb8167]

they will probably drop ObjectiveC as well in a fairly near future.

That seems unlikely given how much of macOS is written in Obj-C.

 

[Joe Dohn]

That's a fairly far fetched conclusion.


It's not. Most people here thought, a year ago, that Rosetta would be discontinued in three to five years: https://www.reddit.com/r/mac/comments/qqntkk
Almost everyone goes in that line, including big websites like AppleInsider.

*

* https://appleinsider.com/inside/ros...ed this problem with,from macOS at some point.
* https://news.ycombinator.com/item?id=25185742

But just because it doesn't put Mac computers in a good light for gaming, it doesn't feel right for you to say that I came with my reasoning out of the blue, because it's not true.

Nothing is permanent. Certainly not with Apple. They deprecate technologies all the time. Like they deprecated Carbon, 32bit and many of their frameworks. Right now they are preparing to deprecate AppKit and UIKit in favour of SwiftUI, and they will probably drop ObjectiveC as well in a fairly near future.

We're just discussing Macs as a gaming platform, and Rosetta is a key piece here. This is obviously why this deprecation is so relevant.

Of course, other technologies being deprecated can make the problem MUCH worse.

 

[Gudi]

Why did Apple switch from RISC-based PowerPC to CISC-based x86?

PowerPC was stagnant, because it wasn't fueled with iPhone money and none of the three AIM alliance members (Apple, IBM, Motorola) could beat Intel on chip expertise. TSMC wasn't ahead of Intel on process nodes. And plugged into the wall PCs were still king. Sometimes it is too zoon for a technology.

 

[Homy]

No. How many recent native games do you think were released?
https://doesitarm.com/games only lists 5.9% games as Apple Silicon native games (no Rosetta / virtualization / Crossover) required.

I counted here, and it only gives me TWELVE games. That's only TWELVE games TWO years after the release of Apple Silicon. That's not even considering some of the games being counted aren't AAA games (e.g, Stardew Valley, Among Us).

It's REALLY bad.

12? Here is a long list of 247 ARM native Mac games. There may be others missing there, like SimCity 4 Deluxe, Grid Legends and No Man's Sky. That's 250. It's not bad. You use a bad source.

 

[jdb8167]

There may be others missing there, like SimCity 4 Deluxe, Grid Legends and No Man's Sky.

As far as I know, No Man's Sky isn't actually out yet. Still waiting.

 

[Homy]

As far as I know, No Man's Sky isn't actually out yet. Still waiting.

Everybody knows that. Your point? The game will be out and it will be native. Are you saying it's not going to be released and be native?

 

[jdb8167]

Everybody knows that. Your point? The game will be out and it will be native. Are you saying it's not going to be released and be native?

It's happened in the past. I'm not much of a gamer but I am looking forward to No Man's Sky. But putting it on a list of existing games for macOS would be misleading.

 

[Joe Dohn]

12? Here is a long list of 247 ARM native Mac games. There may be others missing there, like SimCity 4 Deluxe, Grid Legends and No Man's Sky. That's 250. It's not bad. You use a bad source.

247 is not a "long" list by a long shot.

Also, some of these 247 games are really old. Prince of Persia is a 2008 remake, and Sim City 4 is from 2003. Roller Coaster Tycoon 2 is only available unoffically, and requires extra steps through OpenRCT (open source implementation developed through reverse engineering).

It's nice that these old games are being ported, but none of them are considered modern AAA games, and this list pales in comparison to a console.

 

[Xiao_Xi]

I believe that instead of relying on imprecise terms like "RISC" and "CISC" to characterise architectures, one needs to study the design details and discuss what the instruction sets have in common and how they differ. Blanket statements like "RISC is faster" or "RISC is more power-efficient" are not helpful and only obfuscate the complex reality of CPU design.

The problem with using RISC versus CISC as a lens for comparing modern x86 versus ARM CPUs is that it takes three specific attributes that matter to the x86 versus ARM comparison — process node, microarchitecture, and ISA — crushes them down to one, and then declares ARM superior on the basis of ISA alone.

RISC vs. CISC Is the Wrong Lens for Comparing Modern x86, ARM CPUs - ExtremeTech

Try to investigate the differences between the x86 and ARM processor families (or x86 and the Apple M1), and you'll see the acronyms CISC and RISC. It's a common way to frame the discussion, but not a very useful one. Today, "RISC versus CISC" obscures more than it explains.

www.extremetech.com

 

[leman]

PowerPC was stagnant, because it wasn't fueled with iPhone money and none of the three AIM alliance members (Apple, IBM, Motorola) could beat Intel on chip expertise.

So wait, suddenly it's about money and expertise? I thought it was just "RISC" all the way?

 

[leman]

RISC vs. CISC Is the Wrong Lens for Comparing Modern x86, ARM CPUs - ExtremeTech

Try to investigate the differences between the x86 and ARM processor families (or x86 and the Apple M1), and you'll see the acronyms CISC and RISC. It's a common way to frame the discussion, but not a very useful one. Today, "RISC versus CISC" obscures more than it explains.

www.extremetech.com

It's all nonsense anyway. When someone uses the term "RISC" today they almost always mean load/store architure as opposed to x86's reg/mem. Which is a dumb argument, because:

-- x86
inc     DWORD PTR [eax]

 -- aarch64
ldr     w1, [x0]
add     w1, w1, #1
str     w1, [x0]

Just consider the x86 version to be a shortcut for the three ARM instructions (which is close enough). How does this alone make RISC superior?

As I wrote before, RISC and CISC made sense before because they referred to a specific hardware implementation of the CPU (direct control vs. microcode). Since no modern CPU (aside of the simplest ones) actually works that way anymore, these labels have lost their meaning. Today they merely indicate some sort of vague similarity to the old-school CPU ISA (like, "old RISC was load-store, so if it's not load-store it's not RISC). I have nothing agains nostalgia and vague characterisations, in fact, I believe they can help one to pin down their intuitions, but in a technical discussions one should use precise terms. And not like out friend who can only muster "RISC is a way to design a CPU so that's it's very power efficient".

 

[Homy]

It's happened in the past. I'm not much of a gamer but I am looking forward to No Man's Sky. But putting it on a list of existing games for macOS would be misleading.

247 is not a "long" list by a long shot.

Also, some of these 247 games are really old. Prince of Persia is a 2008 remake, and Sim City 4 is from 2003. Roller Coaster Tycoon 2 is only available unoffically, and requires extra steps through OpenRCT (open source implementation developed through reverse engineering).

It's nice that these old games are being ported, but none of them are considered modern AAA games, and this list pales in comparison to a console.

You are entitled to your opinions but please let’s try not to move the goalposts. Putting announced but unreleased games is not misleading in anyway. Everybody knows NMS isn’t out yet but the question was how many native games are there? The question wasn’t only about existing games but if devs are planning to release native games too. The only reason not putting NMS on the list would be if you know for sure that the game won’t be released or be native and some games not being released in the past is no reason to believe every unreleased game will meet the same fate. Apple has showcased NMS several times since WWDC 2022 in their presentations, on their website and also had special events with reporters where they used the game to show their GPU performance. They showed it again just last week and said it’s coming, meaning they’re following the development closely. Hello games has also announced their future release of the game and collaboration with Apple and had some updates about the work. There is no reason to believe that Apple or HG are not planning to release the game. Hello Games which is a small game dev is not going to make a fool of Apple by not releasing the game.

It’s funny when pc guys here talk about how bad Mac gaming is and how many titles there are for pc they always mention a long list of future pc games in development with a unknown release date in the next 2-3 years. Nobody question them and says it’s misleading but now we have a confirmed game by both Apple and the developer and it doesn’t count? It’s the kind of thing pc guys do to bash Mac gaming.

So first people complain about all the 32-bit games that have disappeared because of Apple. Then they say 64-bit games working perfectly fine in Rosetta don’t count either. Then 12 native games is really bad but when they’re informed that there are actually over 2000% more native games it’s still bad. Then many of them don’t count either because they are old or not AAA. Well, the question wasn’t about AAA games, but how many native AS games there are. You said yourself you weren’t even considering that some of the games weren’t AAA but when I showed there are hundreds of native games, new and old you suddenly change your mind and only count AAA games and start to compare the list with console games? That’s moving goalposts. The point of the list was to show there is noticeable enough interest among devs for porting their old or new games to Apple Silicon. If it wasn’t worth it they wouldn’t bother rewriting those old games. Nobody says Mac gaming couldn’t be better but don’t move goalposts to try to make a point.

 

[Joe Dohn]

So first people complain about all the 32-bit games that have disappeared because of Apple. Then they say 64-bit games working perfectly fine in Rosetta don’t count either.

I think you're mixing up the people here on your explanation, so I'll answer the point addressed at me.

I was pretty clear on what I meant: the games on Rosetta don't count because it is a band-aid. When I run a game on Windows that is considered native, I KNOW that Microsoft will not remove that underlying technology for Windows because of their track record. With many applications, you can actually grab an a Windows 3.0 binary and it will still work UNDER WINDOWS 11. If not out of the box (which sometimes it actually happens), then at least with some hacking.

Now, no one is expecting Apple to deliver 20+ years of backwards compatibility, but we know they tend to discontinue compatibility. Can you GUARANTEE that Rosetta 2 games will be supported 6 or 8 years from now? Apple hasn't made any promises on that.

Then 12 native games is really bad but when they’re informed that there are actually over 2000% more native games it’s still bad. Then many of them don’t count either because they are old or not AAA.

I'm not sure why you want to use a 2000% number to make Apple Silicon look good.
First of all, the percentage is wrong. 252 / 16 = 16 times, or around 1600%.

But even that 1600% is gimmick because most of those 252 are legacy games that HAVE BEEN PORTED BY THE COMMUNITY, as is the case with OpenRCT. This means that there's no commercial interest there – the users have ported it, not the underlying company.

252 is not an impressive number either because even the Switch, which is considered an underpowered console which "only" has Nintendo exclusives actually has 4,469 games to choose from: https://en.wikipedia.org/wiki/List_...xt=There are currently 4469 games,games (Q–Z).


Playstation 5 is less impressive natively, with 546 games listed on the Wiki natively. But it is a very recent console by comparison, and it is compatible with ALL PS4 games, totaling over 4,000 games too. Not only that, but it is compatible with ALL other previous consoles if you use Sony's service, and maybe even more if you use an unofficial emulator.

Well, the question wasn’t about AAA games, but how many native AS games there are. You said yourself you weren’t even considering that some of the games weren’t AAA but when I showed there are hundreds of native games, new and old you suddenly change your mind and only count AAA games and start to compare the list with console games?
That’s moving goalposts.

I haven't moved the goalposts. I argued both about the size catalog of native Apple Silicon games being small / weak AND that it doesn't attract the interest of MODERN AAA game developers. Those are two different, distinct disadvantages.

 

[Homy]

I think you're mixing up the people here on your explanation, so I'll answer the point addressed at me.

No, no mix-up. The first part was directed at the person who talked about it and the last part was directed at the person who talked about it and the middle part was a general comment. The first sentence was for both of you.

I was pretty clear on what I meant: the games on Rosetta don't count because it is a band-aid.

Yes, Rosetta games are not native but it’s not because of developer’s lack of interest, but because Apple did such a good job of creating Rosetta 2 that sometimes it makes very little difference to have a native app. One example is BG3 where the native game and the x86 Rosetta translation many times were equally fast. That’s why many devs haven’t bothered since it would be a waste of time and resources. Last time Apple supported Rosetta for 5.5 years. I wouldn’t be surprised if more devs started to port their games to AS when the time is up. Aspyr for example has already started porting there old catalog as I mentioned before. For them it takes about 4 months for each game. That’s why in my opinion you should count every Rosetta game until they stop working instead of the opposite.

I'm not sure why you want to use a 2000% number to make Apple Silicon look good.
First of all, the percentage is wrong. 252 / 16 = 16 times, or around 1600%.

For the same reason you used 12 instead of 250 to make Apple Silicon look ”really bad”? I wasn’t expecting a math lesson but if you want to be picky we can have one. You’re right about me not being precise but your calculation is actually the wrong one. 2 times more doesn’t mean 200%, it equals 100%. 252 is 1475% larger than 16. I used 250 / 12 = 20.83 or 1983%. Another way is (250-12) / 12 x 100 = 1983%.

But even that 1600% is gimmick because most of those 252 are legacy games that HAVE BEEN PORTED BY THE COMMUNITY, as is the case with OpenRCT. This means that there's no commercial interest there – the users have ported it, not the underlying company.

There are old games on that list ported to AS by the community but I disagree about most of them being so. I just checked Mac Source port and many games there are not on that list. There are Apple Arcade games on the list that work on Mac and that is one benefit of Apple Silicon. Devs can port a game once for all Apple devices, just like NMS coming to iPad and Mac.

252 is not an impressive number either because even the Switch, which is considered an underpowered console which "only" has Nintendo exclusives actually has 4,469 games to choose from: https://en.wikipedia.org/wiki/List_...xt=There are currently 4469 games,games (Q–Z).

If you want to compare to Switch then it’s only fair if you compare the number of all working games on Apple Silicon, including Rosetta games, just as you refer to all currently Nintendo games. There are about 1751 64-bit games for Mac, not 250.

I haven't moved the goalposts. I argued both about the size catalog of native Apple Silicon games being small / weak AND that it doesn't attract the interest of MODERN AAA game developers. Those are two different, distinct disadvantages.

Then we have different definitions of ”moving goalposts”. You used a wrong number of native AS games to describe the situation as ”really bad”. When I provided you with a much larger number you started to compare it to AAA titles and an even larger number from another category and wrote ”this list pales in comparison to a console”. What do console games have to do with native Mac games? If that’s not moving goalposts I can also compare console titles to iPhone/iPad titles and say the profits from console titles pale in comparison to iPhone/iPad titles because Apple earns more than Nintendo, Sony, Microsoft and Activision combined on mobile gaming. Many games like Apex Legends, Call of Duty Warzone and Diablo Immortal work well on Apple Silicon, They just happen to be on iPhone/iPad, not Mac. So the commercial interest for Mac may be weak but for Apple Silicon it’s on a whole another level.

 

[Joe Dohn]

What do console games have to do with native Mac games?

Everything. Many games for PCs are ported to consoles, and many games for consoles are rewritten for PCs. They are not entirely separate categories – especially now, when even the cheapest consoles have the firepower to double as consoles if only Nintendo, Sony or Microsoft would allow them to.

I can also compare console titles to iPhone/iPads titles and say the profits from console titles pale in comparison to iPhone/iPad titles because Apple earns more than Nintendo, Sony, Microsoft and Activision combined on mobile gaming.

As many console gamers would tell you, MOST games on the iPhone are quick cash grabs and pay-to-win. It's either that, or they are cheaper in quality than their console / PC counterparts, with less features.

Just compare Mario Kart Tour with Mario Kart 8.
There's no comparison. Mario Kart Tour is riddled with microtransactions.

But the most aggravating point is that once the game is removed from the App Store in the iPhone, you can't install it on your phone anymore, whereas you absolutely CAN do that on your PC – even games on the Steam store can be backed up and played without Steam, if you ever fear a game will get removed from your library.

That alone is not only cost-saving for users, but helps preserve history. Apple's choice only preserve Apple's wallet.

 

[ksj1]

Either way, I wouldn't count on something I know is temporary.

Almost anything in software is temporary. API's evolve over time, hardware evolves over time.

 

[Gudi]

So wait, suddenly it's about money and expertise? I thought it was just "RISC" all the way?

You never once demonstrated a thought. You literally argued, I haven't heard about it for a couple of decades, therefore it can't exist and be relevant anymore. Microsoft Windows and its "industry standards" are just an economic accident, as it's only a mediocre knockoff of Macintosh OS. Like a Ford Model T its only innovation was to find a way to produce millions of cheap copies of itself. Apple had to give up production in the US to survive and compete with Asian OEMs. Sometimes a cheaper product is all customers are looking for, dragging mankind back into darkness in the process.

 

[MayaUser]

Almost anything in software is temporary. API's evolve over time, hardware evolves over time.

exactly, it seems some here forget about that and talk about the future....in 4-5 years somebody will get to a new platform , get dead, get a new device or get bored by that app or those apps that works today, and the present is all that matter