AMD's Mantle Graphics API For Linux?

[johnc]

Mantle isn't an OpenGL alternative.

Mantle will be a set of frameworks to encapsulate access to OpenGL and OpenCL. Are people truly that dense to think it's a replacement?

So you're saying that Mantle is an API into OpenGL?

You realize Mantle is a low-level console API, right?

[Marc Driftmeyer]

Mantle isn't an OpenGL alternative.

Mantle will be a set of frameworks to encapsulate specific parts of their Hardware via API calls, while in conjunction to interfacing with OpenGL and OpenCL. Are people truly that dense to think it's a replacement? It'll be an API that supports DirectX, OpenGL, OpenCL, not to mention other stuff by accessing the most common APIs without having to directly write strictly in each.

Sorry, but AMD did a truly piss poor job discussing Mantle. I took it as a sound set of APIs to access assembly calls in C/C++ [most likely C++].

AMD is fully committed to OpenGL and OpenCL moving forward. This Mantle opens up access to hardware level optimizations.

[johnc]

Not a lot of people are and that's one of the reasons why. A lot of you are just glossing over the fact that even if it was 100% code compatible with PS4/XB1 (its definitely not), there is still a huge cost to use Mantle. For example, lets say Mantle is 10% faster. I have Direct X and/or OpenGL that already works, why and I going to maintain another API? It costs resources and I'd rather spend it on making the game 3-5% faster for everyone. Other than surfing on the Mantle hype and hoping to get sales from Mantle enthusiasts, there would be no reason if the 10% case were true. Think about the type of games that would benefit from Mantle: BF4, Crysis. Neither of them have Linux Clients. On the other hand if its 20% faster, everyone who loves BF4 would get a GCN card when they upgrade next. GCN market share would rise and other game developers would implement Mantle so their game looks better. My point was if the performances gains are not significant it dies in Windows before they even consider a Linux version.

I'd be surprised if they were overstating the performance gains. What the consoles are able to do with decidedly inferior hardware is pretty amazing, and that's almost all because of the low-level API. I agree that many of the indie devs won't be jumping in feet-first. But the big publishers all run on back-room deals and if AMD can get widespread engine support I can see many games using it. And AMD needs this.

So the danger is that devs work on their PS4/XB1 games (their primary audience)... then drag their feet to port over to Windows. First they'll get Mantle running because that should be a quick port job from the console version. Then they'll put in the work to get their DirectX version working for other architectures. Then they'll say, "Phew! We're exhausted! What else is there left to do?" "Well how about OpenGL?" "OpenGL? Yeah f that... we have no time for that."

[andreano]

simpler driver = less buggy driver

Given that AMD's OpenGL drivers are mediocre, I can see other reasons (than performance) for wanting a way to talk to the hardware at the lowest level ? simpler driver → less buggy driver.

Is Mantle low-level enough that it would make sense to implement something like OpenGL in it? Would such a layered approach lead to less bugs overall?

I also hope Mantle isn't too AMD specific ? the buggiest OpenGL drivers are the blobs of the ARM world (according to the same article), so I guess that's where a low-level API is most acutely needed. Maybe also where it would make most sense in the long run, since ARM is big in the embedded market, where everything is sacrificed for power/price/performance.

[BreezeDM]

hahahahaha oh wow so it's not worth it to mantain a port to mantle since gcn has only 15% of the market? You say this on a linux forum? Let's put aside the whole ease of porting from Xbone and PS4 for a minute and just look at the numbers, Mac OS X + GNU/Linux < 15% market share.

GUYS IT'S OVER NO ONE WILL BE MAKING OPENGL GAMES! SORRY.

Not a lot of people are and that's one of the reasons why. A lot of you are just glossing over the fact that even if it was 100% code compatible with PS4/XB1 (its definitely not), there is still a huge cost to use Mantle. For example, lets say Mantle is 10% faster. I have Direct X and/or OpenGL that already works, why and I going to maintain another API? It costs resources and I'd rather spend it on making the game 3-5% faster for everyone. Other than surfing on the Mantle hype and hoping to get sales from Mantle enthusiasts, there would be no reason if the 10% case were true. Think about the type of games that would benefit from Mantle: BF4, Crysis. Neither of them have Linux Clients. On the other hand if its 20% faster, everyone who loves BF4 would get a GCN card when they upgrade next. GCN market share would rise and other game developers would implement Mantle so their game looks better. My point was if the performances gains are not significant it dies in Windows before they even consider a Linux version.

[johnc]

Not saying it was clearly a shot against Valve's ambitions I'm just saying it's a bit ironic that a user of a system with ~2% market share is saying 15% market share is no big deal.

Ohh I agree with you. Most of the AAA engines are going to have code paths for XB1 and PS4. I think this thing will just make it nearly simple to reuse those same paths on Windows. Frostbite's already got their stuff coded up and apparently the API is available for other engine developers. Sooo... I see this as making it nearly trivial to bring console games over to Windows.

As for Linux......... What's that again?

[cardboard]

If it does get brought over to Linux, could it lead to a better OpenGL implementation than AMD's default?

Rewrite AMD's horrible horrible thing in Mantle...

[chris200x9]

Yes, this was clearly a shot against Valve's ambitions.

Anyone who can't see that is a moron.

Not saying it was clearly a shot against Valve's ambitions I'm just saying it's a bit ironic that a user of a system with ~2% market share is saying 15% market share is no big deal.

Do you realize the amount of work is completely different, right?
Compare supporting OpenGL for Windows, Linux and OS X, where you only need to handle differently context creation. Then, remember there are libraries that handle all of this work for you. Then, take into account how much of the code is actually in the graphics code in a game engine. There's a lot of code there. A big lot. So, porting to Linux is far easier (if you already use OpenGL) than porting from OpenGL to Mantle.

EDIT: Also, you are comparing porting from something that is supported everywhere to something with limited support (you gain performance in 15% of your users, but you don't really get new users) against porting from something less supported to something more supported, which means more users, which means more money.

No tell me about the work involved, I'm genuinely curious. I wouldn't think going from one api to the same api on another platform would be that difficult, wouldn't it just me like going from XNA to iOS game using monogame?

[mrugiero]

hahahahaha oh wow so it's not worth it to mantain a port to mantle since gcn has only 15% of the market? You say this on a linux forum? Let's put aside the whole ease of porting from Xbone and PS4 for a minute and just look at the numbers, Mac OS X + GNU/Linux < 15% market share.

GUYS IT'S OVER NO ONE WILL BE MAKING OPENGL GAMES! SORRY.

Do you realize the amount of work is completely different, right?
Compare supporting OpenGL for Windows, Linux and OS X, where you only need to handle differently context creation. Then, remember there are libraries that handle all of this work for you. Then, take into account how much of the code is actually in the graphics code in a game engine. There's a lot of code there. A big lot. So, porting to Linux is far easier (if you already use OpenGL) than porting from OpenGL to Mantle.

EDIT: Also, you are comparing porting from something that is supported everywhere to something with limited support (you gain performance in 15% of your users, but you don't really get new users) against porting from something less supported to something more supported, which means more users, which means more money.

[johnc]

hahahahaha oh wow so it's not worth it to mantain a port to mantle since gcn has only 15% of the market? You say this on a linux forum? Let's put aside the whole ease of porting from Xbone and PS4 for a minute and just look at the numbers, Mac OS X + GNU/Linux < 15% market share.

GUYS IT'S OVER NO ONE WILL BE MAKING OPENGL GAMES! SORRY.

Yes, this was clearly a shot against Valve's ambitions.

Anyone who can't see that is a moron.