Support NeoGAF

[benny_a]

How many developers are working on DirectX 12 right now? Do we have engines running on it validating MS's claims that the API is just as good as Mantle is in terms of multi core performance and lower overhead?

It would not be the first time a Technology war is won by marketing, F.U.D., PR, and sheer resistance to change. See Intel and their fight against the MS-led ACE consortium seeking to replace the Intel in Wintel, see all the cartel like crap against Rambus in the PC space (sure they were no saint, but their technology did work well and they were essentially forced out by the other DRAM makers).

Sure, DirectX 12 will be on all GPU's and developers will support it, but it looks like we are basically admitting defeat to any chance there will ever be a competitor strong enough to challenge or supplant the current market leader. If this does not make the leader complacent, I do not know what will .

Mantle will be part of a competitor even if Mantle itself loses. Because AMD is giving Mantle tech to Khronos for shaping OpenGL Next.

I'm very disappointed in the Steambox by Valve, because I thought they were going to push in the direction of Linux and get others aboard so OpenGL had a big non-mobile use case again.

 

[Durante]

Personally, as someone who plays at high resolutions / graphics settings with a fast CPU, I'm not all that interested in low-overhead APIs. The good thing about this news from my perspective is that it would seem to confirm future PC releases from Capcom.

Sure, DirectX 12 will be on all GPU's and developers will support it, but it looks like we are basically admitting defeat to any chance there will ever be a competitor strong enough to challenge or supplant the current market leader.

Mantle is not a cross-vendor standard API, and thus cannot supplant or challenge DX. The only thing which could is OpenGL NG (and I'd love for that to happen), but considering Khronos it will probably take forever.

 

[Momentary]

Personally, as someone who plays at high resolutions / graphics settings with a fast CPU, I'm not all that interested in low-overhead APIs. The good thing about this news from my perspective is that it would seem to confirm future PC releases from Capcom.

That's the only thing I thought about as well. A Panta Rhei powered version of DMC 5 on PC to be exact.

 

[Kezen]

It will be interesting to see how relevant Mantle will be when DX12 is around. Will developpers bother by then ?

I have no reason to believe DX12 won't be as fast as Mantle.

 

[Kintaro]

It will be interesting to see how relevant Mantle will be when DX12 is around. Will developpers bother by then ?

I have no reason to believe DX12 won't be as fast as Mantle.

Not sure how relevant Mantle is right now, let alone in the future. Its a non factor.

 

[ashecitism]

I'm very disappointed in the Steambox by Valve, because I thought they were going to push in the direction of Linux and get others aboard so OpenGL had a big non-mobile use case again.

Unfortunately Valve aren't the type of company that push for anything. The most they did was give devs presentations at SDD and then left it up to them to follow or not.

Mantle is not a cross-vendor standard API, and thus cannot supplant or challenge DX. The only thing which could is OpenGL NG (and I'd love for that to happen), but considering Khronos it will probably take forever.

This doesn't bode well for Valve's project then.

 

[TAJ]

Personally, as someone who plays at high resolutions / graphics settings with a fast CPU, I'm not all that interested in low-overhead APIs. The good thing about this news from my perspective is that it would seem to confirm future PC releases from Capcom.

Mantle is not a cross-vendor standard API, and thus cannot supplant or challenge DX. The only thing which could is OpenGL NG (and I'd love for that to happen), but considering Khronos it will probably take forever.

Doesn't SLI increase CPU load? Also, are there CPUs that can maintain 120/144fps in all current high-end games? (given sufficient GPU power)

 

[orava]

Mantle will be part of a competitor even if Mantle itself loses. Because AMD is giving Mantle tech to Khronos for shaping OpenGL Next.

I'm very disappointed in the Steambox by Valve, because I thought they were going to push in the direction of Linux and get others aboard so OpenGL had a big non-mobile use case again.

Well if you look at the situation before they started that project, it's not that bad. Valve has done more to Linux gaming than anyone before this. They have improved the whole public image of Linux significantly. Sure it's not happening overnight but there has been huge improvements as a whole from almost all parties involved. If they can't do it, nobody will.

 

[Xpliskin]

I see what you did there.
That demo scene definetely is the first screenshot of RE7.
Rumor has it umbrella returns.


Topic: Once you try implementing software with asynchronous computing you can't go back. We had our artists get addicted to it because they could preview effects on really large clipmaps while they were being processed, all happening asynchronously(using OpenCL). For real-time purposes the queueing order choices can become quiet tricky

 

[benny_a]

Well if you look at the situation before they started that project, it's not that bad. Valve has done more to Linux gaming than anyone before this. They have improved the whole public image of Linux significantly. Sure it's not happening overnight but there has been huge improvements as a whole from almost all parties involved. If they can't do it, nobody will.

It was maybe phrased too grim. I was merely talking about OpenGL adoption and making it viable. Like a regular platform launch.

What is true that there are developers that now target OpenGL which they haven't done before, but there are some that just bundle a specific wine version.

My intention wasn't to summarize Valve's full effort on that front.

 

[Panajev2001a]

Personally, as someone who plays at high resolutions / graphics settings with a fast CPU, I'm not all that interested in low-overhead APIs. The good thing about this news from my perspective is that it would seem to confirm future PC releases from Capcom.

Mantle is not a cross-vendor standard API, and thus cannot supplant or challenge DX. The only thing which could is OpenGL NG (and I'd love for that to happen), but considering Khronos it will probably take forever.

Hopefully, OpenGL NG becomes a reality then unlike Long Peaks (for some reasons I missed AMD donating Mantle to Open GL NG ...). Still, if AMD maintains its promise to open it up and make it an open cross-vendor API... then it might become a standard too .
Even if AMD created a working group, but remained the ultimate arbiter of the evolution of the specs (hey it works for Google with AOSP and DirectX is certainly cross-vendor, but it is MS who is the ultimate arbiter for the evolution of the API).

 

[orava]

It was maybe phrased too grim. I was merely talking about OpenGL adoption and making it viable. Like a regular platform launch.

What is true that there are developers that now target OpenGL which they haven't done before, but there are some that just bundle a specific wine version.

My intention wasn't to summarize Valve's full effort on that front.

I get you. Also, i quoted the wrong poster.

 

[Kezen]

Hopefully, OpenGL NG becomes a reality then unlike Long Peaks (for some reasons I missed AMD donating Mantle to Open GL NG ...). Still, if AMD maintains its promise to open it up and make it an open cross-vendor API... then it might become a standard too .
Even if AMD created a working group, but remained the ultimate arbiter of the evolution of the specs (hey it works for Google with AOSP and DirectX is certainly cross-vendor, but it is MS who is the ultimate arbiter for the evolution of the API).

Nvidia and AMD always had a heavy hand in graphics APIs, supposedly Mantle is only and only led by AMD.
It's like...Anyone is welcome to board a ship but who's manning it ? Who set its course ?
Yes, only one guy.

 

[Datschge]

Nvidia and AMD always had a heavy hand in graphics APIs, supposedly Mantle is only and only led by AMD.
It's like...Anyone is welcome to board a ship but who's manning it ? Who set its course ?
Yes, only one guy.

That's usually how new projects and features are introduces: The parties convinced that something is of use in the future keep pushing it. Skeptical parties wait until it's proven to do its job and worth their time to implement.

In Mantle's case DICE first convinced AMD to get behind the idea. Since the initial DICE BF4 Mantle reveal Nvidia optimized its DirectX drivers in the affected areas, Microsoft announced DirectX 12 which ought to solve similar issues, OpenGL Next gets AMD's resources to be as efficient as Mantle at what it does, and an increasing number of developers join the Mantle beta as the general idea is that adapting to this not yet finished API will ease them into mitigating to the future DirectX 12 and OpenGL Next.

 

[Durante]

Even if AMD created a working group, but remained the ultimate arbiter of the evolution of the specs

No sane company would ever agree to that, it seems like a death warrant to me. It's like asking AMD or Intel to implement CUDA support.

DirectX is certainly cross-vendor, but it is MS who is the ultimate arbiter for the evolution of the API).

Absolutely, but the thing about MS is that they don't build their own GPUs. As such, there's no conflict of interest.

 

[Stallion Free]

Does this mean they are finally adding exclusive graphical features when using the Mantle renderer?

 

[Panajev2001a]

No sane company would ever agree to that, it seems like a death warrant to me. It's like asking AMD or Intel to implement CUDA support.

Absolutely, but the thing about MS is that they don't build their own GPUs. As such, there's no conflict of interest.

Correct on both front, but in my mind there is always Talisman hehe . Thanks for the corrections (btw, what I read of Talisman was actually very exciting ).