AMD Aims To Give OpenGL A Big Boost, “API Won’t Be The Bottleneck”

[Gloomy]

http://www.dsogaming.com/news/amd-aims-to-give-opengl-a-big-boost-api-wont-be-the-bottleneck/

...AMD caring about OpenGL? Must be a cold day in hell.

 

[Imouto]

AMD aims to outsell Nvidia and fails hardcore.

 

[ShintaiDK]

Shouldnt be hard to give a "big boost" considering how they currently perform in OpenGL.

 

[thilanliyan]

Wow this should make a massive a difference in all those 100s of games that use OGL...oh wait.

I hope devs transition to something other than DX to reduce reliance on MS, but that probably isn't going to happen very soon.

 

[Stuka87]

Wow this should make a massive a difference in all those 100s of games that use OGL...oh wait.

I hope devs transition to something other than DX to reduce reliance on MS, but that probably isn't going to happen very soon.

There actually are a lot (Maybe 100's?) of games that use OpenGL. 99% of them are indy games though. So performance is really not an issue as they are not demanding. Majority of them are on Steam.

 

[nitromullet]

Regardless of how this turns out for AMD, I for one am glad to see some of major players trying to stir things up a bit.

IMO, MS has not leveraged their position as the de facto gamers' OS to the advantage of gamers, or themselves for that matter. They established DX as the dominant API, and then just left gamers hanging while they chase the mobile space, the cloud, or whatever the buzz word of the quarter is.

They have also not created a meaningful synergy between the Xbox 360 and Windows. Third party apps are generally still better than Media Server to stream media from a PC to the Xbox. They could have theoretically created a seamless ecosystem between PC, console, phone, and tablet; but instead they cannibalized the desktop market for mobile. In seven whole years they never gave us a decent Windows app so serve content to the Xbox.

The bottom line is that MS doesn't really care about most of the technology space that interests me the most. So, if there are big players (Valve/AMD) that are willing to challenge the DX monopoly in some sort of fashion, I'm certainly not going to discredit those attempts before seeing what they have to offer.

 

[thilanliyan]

There actually are a lot (Maybe 100's?) of games that use OpenGL. 99% of them are indy games though. So performance is really not an issue as they are not demanding. Majority of them are on Steam.

Fair enough, but how many of them are running dog slow that they need a perf boost? Few to none would be my guess.

 

[ShintaiDK]

Remember OpenGL is also still quite important in the professional segment.

And then there is the SteamOS case, I assume thats why they say it now. While SteamOS needs a minor miracle vs Windows. It is however the only time they get the chance when MS is runnign as they do.

HL3 should also be OpenGL.

 

[postmortemIA]

Remember OpenGL is also still quite important in the professional segment.

And then there is the SteamOS case, I assume thats why they say it now. While SteamOS needs a minor miracle vs Windows. It is however the only time they get the chance when MS is runnign as they do.

HL3 should also be OpenGL.

yes, AMD saw specs on all steam boxes, and nvidia got them all.

 

[NTMBK]

Wow this should make a massive a difference in all those 100s of games that use OGL...oh wait.

I hope devs transition to something other than DX to reduce reliance on MS, but that probably isn't going to happen very soon.

You know, there are one or two applications for graphics cards which don't involve gaming.

 

[thilanliyan]

You know, there are one or two applications for graphics cards which don't involve gaming.

My mistake, but the article talks about gaming, and the devs needing to use the correct extensions for there to be a performance benefit:
"As Sellers claimed, AMD aims to expose all of the hardware of their GPUs with these upcoming high performance extensions of OpenGL, and gamers will be able to get close to theoretical peak and performance. Not only that, but Sellers claimed that games using the modern versions of OpenGL won’t be bottlenecked by the API anymore, meaning that gamers will hit HW limits first."

Either way, better performance (without sacrificing IQ) is definitely good, but currently in the gaming sense better OGL support probably won't affect a lot of people.

 

[tviceman]

Has AMD suddenly hired additional software engineers? They were already ALWAYS playing catch up with functionality and features, now they've added their own proprietary API and are still trying to get CFX working properly at all times. On top of that, they reiterated their support for Linux last week. Sounds like they are making bold claims which will be a long time in coming to fulfill. I think they'd be best off just working on and getting their drivers across all products up to full functioning status and lay off the long-term bold claims.

 

[SiliconWars]

Has AMD suddenly hired additional software engineers? They were already ALWAYS playing catch up with functionality and features, now they've added their own proprietary API and are still trying to get CFX working properly at all times. On top of that, they reiterated their support for Linux last week. Sounds like they are making bold claims which will be a long time in coming to fulfill. I think they'd be best off just working on and getting their drivers across are products up to full functioning status and lay off the long-term bold claims.

It's quite probable that an awful lot of people have been working on Mantle and console API's up until very recently.

 

[toyota]

AMD is always working on something. most of the time it turns out to be nothing great and then we wait for the next thing and then we wait for the next after that. but at least they stay busy. lol

 

[3DVagabond]

Remember OpenGL is also still quite important in the professional segment.

I thought of this as well. Autodesk incorporating Mantle? Others? It's a possibility, anyway.

 

[Abwx]

They were already ALWAYS playing catch up with functionality and features

Of course , with DX11.2 GFX they are running to catch
up with the most advanced nvidia cards that are already DX11.0....

 

[Anarchist420]

The best plugin for dolphin is OpenGL and I'm sure AMD's current lack of IQ in openGL leaves a lot to be desired, so this is a good move.

 

[tviceman]

It's quite probable that an awful lot of people have been working on Mantle and console API's up until very recently.

Nah it was probably the same people that were laid off during their mass firings. Haha but no really, is Mantle the reason they took forever (or still haven't, I don't know) got multi-threaded functionality in their DX11 drivers or that CFX went so long being so buggy or that GCN came out with horribly performance crippled drivers and stayed that way for 9 months?

Lame.

 

[Jodell88]

Has AMD suddenly hired additional software engineers? They were already ALWAYS playing catch up with functionality and features, now they've added their own proprietary API and are still trying to get CFX working properly at all times. On top of that, they reiterated their support for Linux last week. Sounds like they are making bold claims which will be a long time in coming to fulfill. I think they'd be best off just working on and getting their drivers across are products up to full functioning status and lay off the long-term bold claims.

You raise a good point about AMD and software.

The latest beta drivers just introduced OpenGL 4.3 and they've been having problems with their OpenCL compiler almost choking on large kernels (3D renderers). AMD has to learn that no matter how good their hardware may be, if the software is poor then the product is poor.

 

[Red Hawk]

Has AMD suddenly hired additional software engineers? They were already ALWAYS playing catch up with functionality and features, now they've added their own proprietary API and are still trying to get CFX working properly at all times. On top of that, they reiterated their support for Linux last week. Sounds like they are making bold claims which will be a long time in coming to fulfill. I think they'd be best off just working on and getting their drivers across are products up to full functioning status and lay off the long-term bold claims.

Er, always? Remind me which chip maker supported DirectX 10.1, got to DirectX 11 first, released DirectX 11.1 cards first and with the most complete support?

 

[tviceman]

Er, always? Remind me which chip maker supported DirectX 10.1, got to DirectX 11 first, released DirectX 11.1 cards first and with the most complete support?

I thought we were talking about software and drivers, not hardware, but what is the point of being first when it's filled with caveats? Did AMD ever get multi-threaded functionality in their DX11 drivers? That's a pretty big feature to have been left out for so very long.

 

[24601]

Remember OpenGL is also still quite important in the professional segment.

And then there is the SteamOS case, I assume thats why they say it now. While SteamOS needs a minor miracle vs Windows. It is however the only time they get the chance when MS is runnign as they do.

HL3 should also be OpenGL.

Most professional programs are moving to/have moved to DirectX 11 though, so this is an odd time for AMD/ATi to finally care about OpenGL.

 

[24601]

I thought we were talking about software and drivers, not hardware, but what is the point of being first when it's filled with caveats? Did AMD ever get multi-threaded functionality in their DX11 drivers? That's a pretty big feature to have been left out for so very long.

No, they havent.

The most hilarious example is equivalent Nvidia cards are running significantly faster than AMD/ATi cards on AMD processors in Battlefield 4.

AMD/ATi stonewalled DICE into not supporting multi-threaded rendering in Battlefield 3 but it was not able to do so for Battlefield 4.

 

[badb0y]

Most professional programs are moving to/have moved to DirectX 11 though, so this is an odd time for AMD/ATi to finally care about OpenGL.

Steambox?

 

[24601]

Steambox?

AMD/ATi didn't even assign resources to their problems in DirectX in Windows games.

Why the hell would they spend resources to try to fix problems with about as niche an application as it gets?

Also, their work in Windows is not directly applicable in Linux anyways, as they would have to make their driver not ass in Linux before fixing anything wrong with their OpenGL implementation.