In 3 years, 50% of all computers capable of playing AAA games will be Macs

[senttoschool]

It's what studios stated at various conferences. This isn't a secret, this isn't a "studio and I know it, no one else". It was stated at Q&A sessions and panels at the last two GTCs, it was stated at the last few Siggraphs (feel free to contact Carlos Gonzalez Ochoa who is responsible for game related content) and it was stated at GDC to name three conferences where this topic is discussed on a regular basis. These are technology conferences, so you won't find much about specific games and content / IPs there, this is all about technology.

What exactly what stated? Got links also? Did they do deep analysis into the market opportunity of Apple Silicon Macs?

So you have nothing except the old "iOS users spend more than others". Got it. Maybe do the math again not focusing on a single instance as the Apple store, which creates the large chunk of the money with non-AAA games and mostly in-app purchases. When you count other platforms, sum them up per platform, not distribution instance.

You're going off topic. The point is that mobile developers will never leave out iOS for their games because of the market size. Same will be true for MacOS in the future.

Yup, I get it. It's flying over your head. It's ok.

I'd understand you better if you wrote better.

And indeed you're right, numbers are numbers. When it costs a developer more money to port something for a specific platform than what the can make with it, then the numbers show it's a bad idea. And that is the case for complex AAA games. Porting to Mac costs more than sales will get you.

Cool. Now you're finally getting it a little. I agree that porting to macOS is generally not profitable right now.

Porting to Macs will likely be profitable in the future due to Apple Silicon.

Of course it does, but it looks like crap! It doesn't even have proper lighting and shadow system. When you walk around it's a short stop away from static systems. BG3 is an easy to port game, because from a technical point it's trivial. How long have you played it do you really consider the visuals of it as state of the art? Hint, the money they pushed into this for development didn't go into visuals. There's another thread on this forum where I posted one or multiple (can't remember) screenshots of BG3 pointing out the issues with lights and shadows.

It looks like a modern game. It's borderline AAA. It's running smoothly on something as cheap as an M1 Macbook Air. Try running it on an Intel based Macbook Air. Let us know your experience since you're so adamant that Intel iGPUs on Macs can play these types of games already.

 

[diamond.g]

What exactly what stated? Got links also? Did they do deep analysis into the market opportunity of Apple Silicon Macs?


You're going off topic. The point is that mobile developers will never leave out iOS for their games because of the market size. Same will be true for MacOS in the future.


I'd understand you better if you wrote better.



Cool. Now you're finally getting it a little. I agree that porting to macOS is generally not profitable right now.

Porting to Macs will likely be profitable in the future due to Apple Silicon.


It looks like a modern game. It's borderline AAA. It's running smoothly on something as cheap as an M1 Macbook Air. Try running it on an Intel based Macbook Air. Let us know your experience since you're so adamant that Intel iGPUs on Macs can play these types of games already.

I wonder if it will be profitable in three years.

 

[EntropyQ3]

I wonder if it will be profitable in three years.

Nobody knows. Nobody.
* We don't know how much the AS Mac market will grow (but can make a fair guess at this point).
* We don't know what the future AS Mac demographic will look like. (What percentage is interested at all?).
* We don't know how many will be interested in buying games for MacOS (more, since Boot Camp doesn't exist).
* We don't know how the competitive landscape will look (presumably very favourable for the early games).
* We don't know if, given the option, those interested in gaming will choose to buy a game for MacOS over any other platform they have available to them (will they prefer playing Diablo 4 in bed on their MacBook or on their SwitchX?).

Ultimately it will boil down to publishers doing a risk/reward analysis, and then simply make a stab in the dark. Or not.
The data needed to make a good prediction just isn't available.

 

[diamond.g]

Nobody knows. Nobody.
* We don't know how much the AS Mac market will grow (but can make a fair guess at this point).
* We don't know what the future AS Mac demographic will look like. (What percentage is interested at all?).
* We don't know how many will be interested in buying games for MacOS (more, since Boot Camp doesn't exist).
* We don't know how the competitive landscape will look (presumably very favourable for the early games).
* We don't know if, given the option, those interested in gaming will choose to buy a game for MacOS over any other platform they have available to them (will they prefer playing Diablo 4 in bed on their MacBook or on their SwitchX?).

Ultimately it will boil down to publishers doing a risk/reward analysis, and then simply make a stab in the dark. Or not.
The data needed to make a good prediction just isn't available.

I wonder what the breakdown of Mac vs Windows Larian Games sees for BG3.

If the data isn't there, it seems like they will err on the side of just not doing a Mac port, no?

 

[GrumpyCoder]

What exactly what stated? Got links also? Did they do deep analysis into the market opportunity of Apple Silicon Macs?

No intention to port to macOS unless it's a quick port, which is the case for technically simple games. Basic lighting/shadow/visualization. AS interesting, but Metal harder to port to when base is DX12. In general, not worth it, way too small market share. And yes, market opportunity is considered. You'll find links to all recorded sessions and presentations via the usual way, log into the conference portal of the corresponding conference and there's all material available for your level of registration.

You're going off topic. The point is that mobile developers will never leave out iOS for their games because of the market size.

Those are not AAA games however, profit is mostly generated by in-app purchases. They won't leave out any market for these type of games. Think Flappy Bird, if this takes me two weeks with 3-4 hours coding per day, why not release this if it makes me $50k a day? Platform won't matter here.

Same will be true for MacOS in the future.

Uhm, no.

Cool. Now you're finally getting it a little. I agree that porting to macOS is generally not profitable right now.

I got it all along, macOS is not profitable for games, not because of hardware capability, simply because so few people play on macOS.

Porting to Macs will likely be profitable in the future due to Apple Silicon.

For very simple games it is, those are not AAA games however. For technically complex games, very doubtful.

It looks like a modern game. It's borderline AAA.

I think we have very different definitions of what a visually polished game should look like. BG3 looks like crap, sorry.

Look at the lighting and shadows. Sorry, resolution and texture detail aside, this is on a level we've had 15+ years ago. The money for the game did not go into visual representation, which is usually a big chunk of the money. The game is based on Divinity 4, which was adopted to D&D and got a vertical extension. It's still an absolute mess when it comes to large scale UV mapping. That isn't surprising though, the base is still an engine from the Stone Age. But that is to be expected, grab something like freeablo or DGEngine and make the changes required for something like BG3 and you would have exactly the same problems.

I wonder what the breakdown of Mac vs Windows Larian Games sees for BG3.

If the data isn't there, it seems like they will err on the side of just not doing a Mac port, no?

I don't think they'll publish exact numbers. They will probably give you an idea in %.
BG3 is an easy to port game since the engine is simple and was build for that purpose back in the 90s (the original engine that is). It was updated over the years.

The graphics lead engineer behind the engine, gave a talk back in 2019 about porting games to Metal, which outlines their approach, including some approaches like PBR, light probes and clustered lighting. In general, they focus on porting from DX11 to Metal 1.2, keeping things simple. That also means you won't have any luck supporting "modern" binding tiers with this approach. Bad for cutting edge AAA titles, good for stuff that has little technical features and looks "simple". The approach they take requires very, very little of the code base to be changed in order to port to macOS, IIRC it's under 3% of the total project. So in that sense, it totally makes sense as they don't have to worry about things that other AAA developers, with more polished titles, are facing.

On a side note, these guys are also working for Elverils, focusing on security analysis, porting code and such. So much of where this is coming from is paid for by someone else and then flows back into their games. This is similar to Laminar Research primarily working in research and releasing X-Plane as a result of the research work.

 

[diamond.g]

No intention to port to macOS unless it's a quick port, which is the case for technically simple games. Basic lighting/shadow/visualization. AS interesting, but Metal harder to port to when base is DX12. In general, not worth it, way too small market share. And yes, market opportunity is considered. You'll find links to all recorded sessions and presentations via the usual way, log into the conference portal of the corresponding conference and there's all material available for your level of registration.

Those are not AAA games however, profit is mostly generated by in-app purchases. They won't leave out any market for these type of games. Think Flappy Bird, if this takes me two weeks with 3-4 hours coding per day, why not release this if it makes me $50k a day? Platform won't matter here.

Uhm, no.

I got it all along, macOS is not profitable for games, not because of hardware capability, simply because so few people play on macOS.

For very simple games it is, those are not AAA games however. For technically complex games, very doubtful.

I think we have very different definitions of what a visually polished game should look like. BG3 looks like crap, sorry.

Look at the lighting and shadows. Sorry, resolution and texture detail aside, this is on a level we've had 15+ years ago. The money for the game did not go into visual representation, which is usually a big chunk of the money. The game is based on Divinity 4, which was adopted to D&D and got a vertical extension. It's still an absolute mess when it comes to large scale UV mapping. That isn't surprising though, the base is still an engine from the Stone Age. But that is to be expected, grab something like freeablo or DGEngine and make the changes required for something like BG3 and you would have exactly the same problems.


I don't think they'll publish exact numbers. They will probably give you an idea in %.
BG3 is an easy to port game since the engine is simple and was build for that purpose back in the 90s (the original engine that is). It was updated over the years.

The graphics lead engineer behind the engine, gave a talk back in 2019 about porting games to Metal, which outlines their approach, including some approaches like PBR, light probes and clustered lighting. In general, they focus on porting from DX11 to Metal 1.2, keeping things simple. That also means you won't have any luck supporting "modern" binding tiers with this approach. Bad for cutting edge AAA titles, good for stuff that has little technical features and looks "simple". The approach they take requires very, very little of the code base to be changed in order to port to macOS, IIRC it's under 3% of the total project. So in that sense, it totally makes sense as they don't have to worry about things that other AAA developers, with more polished titles, are facing.

On a side note, these guys are also working for Elverils, focusing on security analysis, porting code and such. So much of where this is coming from is paid for by someone else and then flows back into their games. This is similar to Laminar Research primarily working in research and releasing X-Plane as a result of the research work.

Do you happen to have any insight into how they were able to get FSR working in Metal?

 

[GrumpyCoder]

Do you happen to have any insight into how they were able to get FSR working in Metal?

No I don't, if I'd have to guess I'd say they didn't. FSR is available (https://github.com/GPUOpen-Effects), it's known how it works, so my guess is they did their own version. Another option would be via a translation layer since FSR works on Vulkan, but I doubt that as it would add a significant overhead.

 

[leman]

No I don't, if I'd have to guess I'd say they didn't. FSR is available (https://github.com/GPUOpen-Effects), it's known how it works, so my guess is they did their own version. Another option would be via a translation layer since FSR works on Vulkan, but I doubt that as it would add a significant overhead.

The FSR shader code seems to be plain enough so that it compiles for most shading languages, so fixing it for Metal is probably fairly trivial. You just need to replace the inout etc. with references and it should work.

 

[diamond.g]

The FSR shader code seems to be plain enough so that it compiles for most shading languages, so fixing it for Metal is probably fairly trivial. You just need to replace the inout etc. with references and it should work.

I wonder what took them so long in getting patch 6 working on metal then.

 

[leman]

I wonder what took them so long in getting patch 6 working on metal then.

It worked on metal pretty much on release, it's the M1 version that took longer. Talking completely out of my ass here, my bet is either on some sort of subtle hard to pin bug OR implementing additional optimizations for Apple Silicon.

 

[diamond.g]

It worked on metal pretty much on release, it's the M1 version that took longer. Talking completely out of my ass here, my bet is either on some sort of subtle hard to pin bug OR implementing additional optimizations for Apple Silicon.

Ah, so they have to program 2 paths for Metal?

 

[diamond.g]

No intention to port to macOS unless it's a quick port, which is the case for technically simple games. Basic lighting/shadow/visualization. AS interesting, but Metal harder to port to when base is DX12. In general, not worth it, way too small market share. And yes, market opportunity is considered. You'll find links to all recorded sessions and presentations via the usual way, log into the conference portal of the corresponding conference and there's all material available for your level of registration.

Those are not AAA games however, profit is mostly generated by in-app purchases. They won't leave out any market for these type of games. Think Flappy Bird, if this takes me two weeks with 3-4 hours coding per day, why not release this if it makes me $50k a day? Platform won't matter here.

Uhm, no.

I got it all along, macOS is not profitable for games, not because of hardware capability, simply because so few people play on macOS.

For very simple games it is, those are not AAA games however. For technically complex games, very doubtful.

I think we have very different definitions of what a visually polished game should look like. BG3 looks like crap, sorry.

Look at the lighting and shadows. Sorry, resolution and texture detail aside, this is on a level we've had 15+ years ago. The money for the game did not go into visual representation, which is usually a big chunk of the money. The game is based on Divinity 4, which was adopted to D&D and got a vertical extension. It's still an absolute mess when it comes to large scale UV mapping. That isn't surprising though, the base is still an engine from the Stone Age. But that is to be expected, grab something like freeablo or DGEngine and make the changes required for something like BG3 and you would have exactly the same problems.


I don't think they'll publish exact numbers. They will probably give you an idea in %.
BG3 is an easy to port game since the engine is simple and was build for that purpose back in the 90s (the original engine that is). It was updated over the years.

The graphics lead engineer behind the engine, gave a talk back in 2019 about porting games to Metal, which outlines their approach, including some approaches like PBR, light probes and clustered lighting. In general, they focus on porting from DX11 to Metal 1.2, keeping things simple. That also means you won't have any luck supporting "modern" binding tiers with this approach. Bad for cutting edge AAA titles, good for stuff that has little technical features and looks "simple". The approach they take requires very, very little of the code base to be changed in order to port to macOS, IIRC it's under 3% of the total project. So in that sense, it totally makes sense as they don't have to worry about things that other AAA developers, with more polished titles, are facing.

On a side note, these guys are also working for Elverils, focusing on security analysis, porting code and such. So much of where this is coming from is paid for by someone else and then flows back into their games. This is similar to Laminar Research primarily working in research and releasing X-Plane as a result of the research work.

WRT the shadows, to be fair most games are littered with non-shadow casting light sources (World of Warcraft is an offender as well).

 

[leman]

Ah, so they have to program 2 paths for Metal?

It is not that common to have multiple paths anyway (like one for Nvidia and one for AMD, or different workarounds for specific driver or GPU versions). But again, please don't forget that I was widely speculating, I don't know anything of relevance.

 

[senttoschool]

No intention to port to macOS unless it's a quick port, which is the case for technically simple games. Basic lighting/shadow/visualization. AS interesting, but Metal harder to port to when base is DX12. In general, not worth it, way too small market share. And yes, market opportunity is considered. You'll find links to all recorded sessions and presentations via the usual way, log into the conference portal of the corresponding conference and there's all material available for your level of registration.

Those are not AAA games however, profit is mostly generated by in-app purchases. They won't leave out any market for these type of games. Think Flappy Bird, if this takes me two weeks with 3-4 hours coding per day, why not release this if it makes me $50k a day? Platform won't matter here.

Uhm, no.

I got it all along, macOS is not profitable for games, not because of hardware capability, simply because so few people play on macOS.

For very simple games it is, those are not AAA games however. For technically complex games, very doubtful.

I think we have very different definitions of what a visually polished game should look like. BG3 looks like crap, sorry.

Look at the lighting and shadows. Sorry, resolution and texture detail aside, this is on a level we've had 15+ years ago. The money for the game did not go into visual representation, which is usually a big chunk of the money. The game is based on Divinity 4, which was adopted to D&D and got a vertical extension. It's still an absolute mess when it comes to large scale UV mapping. That isn't surprising though, the base is still an engine from the Stone Age. But that is to be expected, grab something like freeablo or DGEngine and make the changes required for something like BG3 and you would have exactly the same problems.


I don't think they'll publish exact numbers. They will probably give you an idea in %.
BG3 is an easy to port game since the engine is simple and was build for that purpose back in the 90s (the original engine that is). It was updated over the years.

The graphics lead engineer behind the engine, gave a talk back in 2019 about porting games to Metal, which outlines their approach, including some approaches like PBR, light probes and clustered lighting. In general, they focus on porting from DX11 to Metal 1.2, keeping things simple. That also means you won't have any luck supporting "modern" binding tiers with this approach. Bad for cutting edge AAA titles, good for stuff that has little technical features and looks "simple". The approach they take requires very, very little of the code base to be changed in order to port to macOS, IIRC it's under 3% of the total project. So in that sense, it totally makes sense as they don't have to worry about things that other AAA developers, with more polished titles, are facing.

On a side note, these guys are also working for Elverils, focusing on security analysis, porting code and such. So much of where this is coming from is paid for by someone else and then flows back into their games. This is similar to Laminar Research primarily working in research and releasing X-Plane as a result of the research work.

I'm sorry but I didn't read it.

I actually don't know what your stance is on this entire thread. Maybe you can start with that? Like what are you actually arguing for?

 

[diamond.g]

I'm sorry but I didn't read it.

I actually don't know what your stance is on this entire thread. Maybe you can start with that? Like what are you actually arguing for?

I dont want to put words in his mouth, but I think he is saying that big studios don’t plan on writing games for Metal/M1/macOS because there is no ROI. Especially studio’s that write to DX12/U, unless I misunderstand, there is no 1 to 1 for DX12/U API calls for Metal and thus some games would have to be rewritten from scratch to work.

That these 27 million M1 Mac users are not all buying games or are expected to buy games like the 30 something million gaming PC folk since the gaming PC folk are not 100% of the PC market like the Mac users are for the Mac market.

 

[quarkysg]

I dont want to put words in his mouth, but I think he is saying that big studios don’t plan on writing games for Metal/M1/macOS because there is no ROI. Especially studio’s that write to DX12/U, unless I misunderstand, there is no 1 to 1 for DX12/U API calls for Metal and thus some games would have to be rewritten from scratch to work.

That these 27 million M1 Mac users are not all buying games or are expected to buy games like the 30 something million gaming PC folk since the gaming PC folk are not 100% of the PC market like the Mac users are for the Mac market.

I do not have data to back up my thoughts, but if we think about it, we can group Mac users into 2 main groups:

1. School going kids
2. Professionals

The first group typically uses the lower range Macs, and these Macs are mainly using Intel iGPUs that will not be suitable for gaming that will attract kids in any serious manner. This group of users play a lot of games, and many grew up with iOS devices.

The second group mainly uses their Macs for work, and wouldn't be running games on their work Mac.

Now, if the Macs used by the first group is good enough to play AAA games (whatever that means), it will tempt developers to develop for it. I would think APIs for the platform is a minor hurdle if the incentive is there. If the Mac market suddenly grew 10 fold, games developers will rush to release for the Mac platform even if Metal is 20x more difficult to code for compared to say DX12.

 

[diamond.g]

I do not have data to back up my thoughts, but if we think about it, we can group Mac users into 2 main groups:

1. School going kids
2. Professionals

The first group typically uses the lower range Macs, and these Macs are mainly using Intel iGPUs that will not be suitable for gaming that will attract kids in any serious manner. This group of users play a lot of games, and many grew up with iOS devices.

The second group mainly uses their Macs for work, and wouldn't be running games on their work Mac.

Now, if the Macs used by the first group is good enough to play AAA games (whatever that means), it will tempt developers to develop for it. I would think APIs for the platform is a minor hurdle if the incentive is there. If the Mac market suddenly grew 10 fold, games developers will rush to release for the Mac platform even if Metal is 20x more difficult to code for compared to say DX12.

Is the thought that the Mac Gamer market is between 10% and 100% of the total Mac market?

We know PC gaming is roughly 10% of the whole PC market, but the thread premise is that AAA games don’t run on anything less than a GTX1060 because that is the single most used GPU according to Steam. Even though most AAA games will literally run on PC iGPU’s (there are some that actually look at the hardware the game is ran on and prevents the game from starting, but that isn’t super common).

 

[leman]

I do not have data to back up my thoughts, but if we think about it, we can group Mac users into 2 main groups:

1. School going kids
2. Professionals

The first group typically uses the lower range Macs, and these Macs are mainly using Intel iGPUs that will not be suitable for gaming that will attract kids in any serious manner. This group of users play a lot of games, and many grew up with iOS devices.

The second group mainly uses their Macs for work, and wouldn't be running games on their work Mac.

Now, if the Macs used by the first group is good enough to play AAA games (whatever that means), it will tempt developers to develop for it. I would think APIs for the platform is a minor hurdle if the incentive is there.

I see this similarly, and given the fact that Apple mentions gaming in connection with the cheaper Macs — but not the "pro" Macs, leads me to suspect that Apple's marketing department also sees it this way.

One problem with "gaming" is that it can be a very dysfunctional and toxic culture. The paradox of gaming is that it is fundamentally elitist (in hardware requirements, "accepted" skill, and time investment), yet virtually none of the people playing games belong to the elite. It is unlikely that entry-level Apple-Silicon Mac will ever be able to compete performance-wise with high-end gaming PCs, even if they are more than capable enough to run contemporary games. This obviously creates a cultural hurdle that can be either broken with large targeted investments (as many here suggest), or by, well, letting the passage of time and natural cultural shift do it's thing.

 

[diamond.g]

Looks like even Feral Interactive doesn’t want to port a game they got working on the M1 iPad with keyboard and mouse to Apple Silicon macOS. That can’t be a good look…

 

[GrumpyCoder]

I dont want to put words in his mouth, but I think he is saying that big studios don’t plan on writing games for Metal/M1/macOS because there is no ROI. Especially studio’s that write to DX12/U, unless I misunderstand, there is no 1 to 1 for DX12/U API calls for Metal and thus some games would have to be rewritten from scratch to work.

Spot on, yes. Thank you. DX12->Metal, depending on features maybe not everything from scratch, but at least with a significant amount of work. It really depends on the game.

Now, if the Macs used by the first group is good enough to play AAA games (whatever that means), it will tempt developers to develop for it.

Depends on the game. For technically simple games, sure. Most AAA games are not simple, but even if your first group is good enough for games, there's still not enough potential buyers for expensive ports of games. How many group #1 users are there vs group #2 users? Mac are mainly used for work (as are PCs), many running Windows on it too for business software. Then the occasional gaming. Not every Mac owner is interested in playing a game, maybe some chess or poker aside.

Looks like even Feral Interactive doesn’t want to port a game they got working on the M1 iPad with keyboard and mouse to Apple Silicon macOS. That can’t be a good look…

Isn't this out already for macOS? I mean, it's an old game that was released back in 2014/2015 for computers/consoles and only getting now a mobile Android/iOS port now?

 

[diamond.g]

Spot on, yes. Thank you. DX12->Metal, depending on features maybe not everything from scratch, but at least with a significant amount of work. It really depends on the game.

Depends on the game. For technically simple games, sure. Most AAA games are not simple, but even if your first group is good enough for games, there's still not enough potential buyers for expensive ports of games. How many group #1 users are there vs group #2 users? Mac are mainly used for work (as are PCs), many running Windows on it too for business software. Then the occasional gaming. Not every Mac owner is interested in playing a game, maybe some chess or poker aside.

Isn't this out already for macOS? I mean, it's an old game that was released back in 2014/2015 for computers/consoles and only getting now a mobile Android/iOS port now?

It isn’t Apple Silicon native as far as I could tell.

 

[leman]

It isn’t Apple Silicon native as far as I could tell.

Why invest resources into making an old game native if it already works under Rosetta?

 

[diamond.g]

Why invest resources into making an old game native if it already works under Rosetta?

The YT video implied that the older version was missing some newer sharpening features and effects. Is that not the case?

 

[leman]

The YT video implied that the older version was missing some newer sharpening features and effects. Is that not the case?

I have no idea. Never knew this game existed before you mentioned it. I’m just wondering whether investing into implementing new features for a game that was released five years ago makes any business sense. How are you going to make the money back? Everyone who wanted to buy the game likely already did.

 

[diamond.g]

I have no idea. Never knew this game existed before you mentioned it. I’m just wondering whether investing into implementing new features for a game that was released five years ago makes any business sense. How are you going to make the money back? Everyone who wanted to buy the game likely already did.

I mean to be fair they didn’t make a aTV version either so I guess they are consistent with only supporting money making platforms…

(I hadn’t heard much about this game recently either, I actually forgot it was a cross generation PS3-PS4 game)