thats good news , i wonder if they will make it work on cards which are not supported right now.

The Khronos group will obviously not allow OpenGL to end up being GCN-only. That would defy the open nature of Linux. AMD is just letting them learn from Mantle and use its code so they can make the new OpenGL as good as possible. AMD is being charitable unlike the selfish, corporate overrun companies like MS and Nvidia.

Uh if the Mantle additions are implemented via extensions then it definitely can be hardware specific. There are plenty of things vendor specific now in OpenGL. Also OpenGL already has low level "close to metal" support via that extension system, it's what makes OpenGL more powerful than DirectX. It also already has extremely low CPU overhead. So I can't really see what Mantle is bringing to the table unless it's hardware specific. And you're second statement is just wrong. Nvidia contributed more code to linux in 2013 then AMD did and Microsoft was a top contributor for years. Plus Nvidia open sourced it's K1 graphics support + is now contributing documentation for the Nouveau project. Both AMD and Nvidia are charitable when it suits them. Stop being a fanboy.

Uh if the Mantle additions are implemented via extensions then it definitely can be hardware specific. There are plenty of things vendor specific now in OpenGL. Also OpenGL already has low level "close to metal" support via that extension system, it's what makes OpenGL more powerful than DirectX. It also already has extremely low CPU overhead. So I can't really see what Mantle is bringing to the table unless it's hardware specific. And you're second statement is just wrong. Nvidia contributed more code to linux in 2013 then AMD did and Microsoft was a top contributor for years. Plus Nvidia open sourced it's K1 graphics support + is now contributing documentation for the Nouveau project. Both AMD and Nvidia are charitable when it suits them. Stop being a fanboy.

When i aggree with you, today many OpenGL developpers are alarmed by the work have done Nvidia on OpenGL so far. theres a reason why the name "Green Mafia" is given to some developpers group. ( theres some nice article to read on OpenGL / linux side who have been done this year on it ). Now back to the Mantle question.. The point is to let OpenGL take everything they want from Mantle and help them to code it.. If Mantle user specific things to GCN today, this is too because it use some new way of pin the informations, use them, who will not be possible on old hardware. ( because they miss this technology, as they will not been able to do tiles ressource etc etc )... Its not a problem for Nvidia gpu because all of this is too part of last technology implemented for example in DX11.. Dx11.2 or part of latest developpement on GPU hardware.. Ofc the way is coded Mantle today will surely not be 100% compatible with Nvidia hardware, but with instruction made for Nvidia hardware this should be really easy to implement too on Intel and Nvidia gpu.. ( Intel is allready studying Mantle approach for their own IGP). I dont think it is a problem for OpenGL, OpenGL next will stay backward compatible with older iteration, but just hardware wise, you will not get all the benefit of it... As Mantle or DX, for OpenGL next, if you want to get the most of latest hardware, theres no miracle, you need to use the latest technology, features of thoses hardwares. Otherwise, you can still find another way, but it will never been as efficient than by use the hardware possibilty.. for an API who want to use new hardware the most efficient way, theres no other possibility.

Well i am surprised, since mantle still is closed...

Actually AMD says anyone who wants to use the Mantel API can do so, even nVidia.

Actually AMD says anyone who wants to use the Mantel API can do so, even nVidia.

Uh yeah they said that but it kinda requires them to release the source which they haven't done yet. Intel even requested access but they denied them. http://techreport.com/news/26682/intel-asked-amd-for-mantle-api-spec

Uh yeah they said that but it kinda requires them to release the source which they haven't done yet. Intel even requested access but they denied them. /QUOTE] From the source: "Intel's request was rebuffed, though. Mantle is still in a closed beta, and AMD doesn't plan on opening the API until version 1.0 is complete. That public release, which Huddy said will be available to Intel and "anyone who wants to participate," is expected to be available before the end of the year." Only because it's in beta. They'll get access once release 1.0 hits.

From the source: "Intel's request was rebuffed, though. Mantle is still in a closed beta, and AMD doesn't plan on opening the API until version 1.0 is complete. That public release, which Huddy said will be available to Intel and "anyone who wants to participate," is expected to be available before the end of the year." Only because it's in beta. They'll get access once release 1.0 hits.

You mean after they finish polishing it for GCN. How altruistic of them. I'm sorry, but I really disrespect amd for poor marketing techniques and I think this is just one of them. They seem to be the altruistic guys for giving mantle away when in fact, I don't think they'll be giving anything away. Mantle will be optimized for GCN, it makes perfect sense from a business point of view. And the others will have to follow suit and further extend mantle (a different environment) for their own architectures? When dx12 is around the corner? And while opengl is gaining popularity? It just doesn't seem likely. That's my 2 cents.

I can see this as great news for AMD users as well as Nvidia users. AMD users might get improved OpenGL support something they been needing like forever and this might Nvidia get access to Mantle.

AMD is being charitable unlike the selfish, corporate overrun companies like MS and Nvidia.

Not to burst your bubble, but "charity" has nothing to do with it. AMD's sole purpose is to serve its shareholders (not the customers; we're just the means to an end) so while money might not exchange hands (directly) you better believe it that it's a business decision for business reasons. There's absolutely nothing selfish about how Intel or NVIDIA do business, they do, after-all, give us top tiered technology, and at least we know where they stand, whereas AMD, at least with Mantle, kept going back and forth about how it's open source to only works on GCN to now this. Charitable...ha!

You mean after they finish polishing it for GCN. How altruistic of them. I'm sorry, but I really disrespect amd for poor marketing techniques and I think this is just one of them. They seem to be the altruistic guys for giving mantle away when in fact, I don't think they'll be giving anything away. Mantle will be optimized for GCN, it makes perfect sense from a business point of view. And the others will have to follow suit and further extend mantle (a different environment) for their own architectures? When dx12 is around the corner? And while opengl is gaining popularity? It just doesn't seem likely. That's my 2 cents.

Why would you release something thats unfinished ? If they say publicly that they are giving Mantle to OpenGL group for OpenGL next, I say props to them. Thats great news. I'll hold them to that. A lot of people would like OpenGL to retake the crown for Dev purposes, since it is soo cross platform, and can even be easily ported to mobile devices. (And Linux). I'm personnally really hoping that soon I'll be able to ditch Windows for Gaming alltogether. (not gonna happen, but i can always wish). 🙂

Why would you release something thats unfinished ? If they say publicly that they are giving Mantle to OpenGL group for OpenGL next, I say props to them. Thats great news. I'll hold them to that. A lot of people would like OpenGL to retake the crown for Dev purposes, since it is soo cross platform, and can even be easily ported to mobile devices. (And Linux). I'm personnally really hoping that soon I'll be able to ditch Windows for Gaming alltogether. (not gonna happen, but i can always wish). 🙂

Agreed with that, porting will be a hell of a lot easier when that happens.

You mean after they finish polishing it for GCN. How altruistic of them. I'm sorry, but I really disrespect amd for poor marketing techniques and I think this is just one of them. They seem to be the altruistic guys for giving mantle away when in fact, I don't think they'll be giving anything away. Mantle will be optimized for GCN, it makes perfect sense from a business point of view. And the others will have to follow suit and further extend mantle (a different environment) for their own architectures? When dx12 is around the corner? And while opengl is gaining popularity? It just doesn't seem likely. That's my 2 cents.

I do not agree with you xIcarus, Mantle is still new to the market, it's buggy (I use it in BF4 and sometimes swear at it) and AMD needs to put much more work and effort to get it running at it's optimum performance. AMD says that it will run on any other architecture than GCN, however I can understand that they want to finish their product before letting other companies use it, which is what any company launching new product could want. AMD stands for open standards (i will mention OpenCL or FreeSync as best examples), contrary to nVidia's PhysX. Personally I do not think that Mantle has a chance competing against DX12 so it's nice to see that they are willing to collaborate with OpenGL to make the standard widely accepted and more popular. Again, by doing this it's unlikely that they will sell more GPU's as OpenGL does not have enough of an impact in consumer space, but it means that MS cannot afford not to upgrade their DX platform, and ultimately, progress is best for all of us!

Uh yeah they said that but it kinda requires them to release the source which they haven't done yet. Intel even requested access but they denied them. http://techreport.com/news/26682/intel-asked-amd-for-mantle-api-spec

They will not publish the SDK publicly before the v 1.0 is out.. But you can imagine that when developpers are including right now Mantle in Crytek, Frostbyte or specific games, they have the source, SDK. lol.

They will not publish the SDK publicly before the v 1.0 is out.. But you can imagine that when developpers are including right now Mantle in Crytek, Frostbyte or specific games, they have the source, SDK. lol.

They might have the SDK, I doubt they have the source.

I do not agree with you xIcarus, Mantle is still new to the market, it's buggy (I use it in BF4 and sometimes swear at it) and AMD needs to put much more work and effort to get it running at it's optimum performance. AMD says that it will run on any other architecture than GCN, however I can understand that they want to finish their product before letting other companies use it, which is what any company launching new product could want. AMD stands for open standards (i will mention OpenCL or FreeSync as best examples), contrary to nVidia's PhysX. Personally I do not think that Mantle has a chance competing against DX12 so it's nice to see that they are willing to collaborate with OpenGL to make the standard widely accepted and more popular. Again, by doing this it's unlikely that they will sell more GPU's as OpenGL does not have enough of an impact in consumer space, but it means that MS cannot afford not to upgrade their DX platform, and ultimately, progress is best for all of us!

I agree with you on some points, but I disagree on others. Lets see: Mantle is new and buggy, that's why it would have probably been a wise decision to release it and work on it with the community and with other companies interested in it (like intel). AMD at this point is going to hold their development team on mantle (which is not that talented, look at amd's drivers), and delaying full deployment in order to have an edge until others catch up in using mantle. The problem I see: at the time mantle is out of beta, dx12 will likely be very close. And if dx12 delivers, mantle will be forgotten in large part. And opengl already has that close to the metal access that mantle strives to offer. Khronos group will most likely not use that much code from the mantle api when making opengl next. That's my 2 cents. I seriously think this is just another marketing strategy from amd's part, and if it really is, it's damn disgusting. Freesync, yes, I agree. I REALLY hope that becomes a thing. But what's up with opencl? They did not start the opencl initiative, iirc apple started the whole thing and pushed it to khronos group (they're a part of this consortium aswell). And I also agree with nvidia physx. I have no clue why the bastards are still keeping it closed-source, it doesn't make any sense and it stains their already sinking reputation. But I understood something else from this article. They have made mantle available to opengl. That will simply mean that opengl will have gcn optimizations in place IF they will indeed be pushed forward. The point of this is that amd was always behind in opengl implementations. The alternative for dx on linux is opengl. And if linux gaming becomes popular, amd will have a huge problem because their cards will not be able to compete in the opengl environment. Remember when I said a few lines before that this might be a marketing strategy? Amd are not trying to be selfless, they are trying to save their asses. This is what it sounds like to me. Now I may be wrong, but this is my opinion. We will see in the coming future what will happen. ps. keep in mind, I'm not hating on amd. I'm hating on mantle and on false advertising. It's always seemed to me that mantle had no place in the world of video APIs, where opengl is alive and becoming a threat to direct3d. It all depends on if direct3d 12 will deliver. And it probably will, it's been in development for a good amount of time. And I dislike false advertising, just like nvidia said a long time ago that a 8600gt could run dx10 games decently which was absolutely not true.

I like this whole mantle idea just because it can be run on any platform microsoft cant force us to buy a new revision of windows for every version of dx i mean they still will but it puts pressure on them. Mantle may not have a place for long but it pushes the others forward.

I think everybody is missing the point about AMD offering the design of the Mantle API to Khronos as a blueprint to OpenGL NG. If Khronos uses ideas from Mantle, it doesn't mean OpenGL NG will be optimized for GCN, it just means its using design ideas. The Mantle API itself, while it is meant for GCN, doesn't necessarily have to be specific to AMD hardware. If you look at Apple's Metal API, its very similar. Its just a collection of data structures and methods for grouping data and commands for submission to the GPU in the most optimal way.