[PcPer] Some mantle news from GDC

[n0x1ous]

http://www.pcper.com/news/Graphics-Cards/GDC-15-AMD-Mantle-Might-Be-Dead-We-Know-It-No-Public-SDK-Planned

Sounds like it's getting pushed to the backburner as expected.

 

[Noctifer616]

http://community.amd.com/community/...g/2015/03/02/on-apis-and-the-future-of-mantle

A link to the the AMD blog post.

 

[ShintaiDK]

So Mantle is officially dead. DX12 won on TKO.

 

[sontin]

And no free SDK. It's quite hilarious when you review the last 12 months and how AMD is all about openness.

From February 2014:

AMD told TechRadar that it'd be willing to make Mantle, or an API based on it, available across the industry. Even, the company said, if it means Mantle is adopted by competitors like Nvidia.
"Mantle for now is straight up in a closed ecosystem, a closed beta, which you have to do in a complicated project like this to get it off the ground. It's us and a few key game developers," Robert Hallock, technical communications, AMD Graphics & Gaming, told us in a recent interview.
"After that phase is done, we do hope that Mantle becomes an industry standard. We'll be releasing a public SDK later this year, and hope that others adopt it. If they don't adopt it itself, then we hope they adopt APIs similar to it that become an industry standard for PC gaming."

http://www.techradar.com/news/compu...g-api-to-become-the-industry-standard-1218560

 

[caswow]

thanks amd for revolutionizing the industry :thumbsup:

 

[Rvenger]

There was an interview on youtube with Richard Huddy explaining the possibility of Mantle being dropped. Also, I believe PCPer is incorrect. Mantle will not be used on consoles, why would that even be a thought to use a console like based API on a console?

 

[96Firebird]

So long ol' friend, we hardly knew ya.

 

[Bubbleawsome]

I was hoping we would see something come of this.

 

[Paul98]

Sounds to me like Mantle "1.0" capabilities will be in DX12.

I'll be interested to see what else we learn thursday.

 

[JimKiler]

Sounds to me like Mantle "1.0" capabilities will be in DX12.

I'll be interested to see what else we learn thursday.

And it is a win if those advances are in DX12.

Tangent question, I remember hearing Nvidia's mantle rebuttal stating they could do something similar. Did that ever materialize in their drivers or game coding?

 

[n0x1ous]

And it is a win if those advances are in DX12.

Tangent question, I remember hearing Nvidia's mantle rebuttal stating they could do something similar. Did that ever materialize in their drivers or game coding?

Well there Dx11 drivers are miles ahead of amds with respect to cpu usage and multithreading

 

[ShintaiDK]

Well there Dx11 drivers are miles ahead of amds with respect to cpu usage and multithreading

And DX12 from the looks of it.

 

[sontin]

Nope, just better hardware with DX12.
nVidia's command processor can process much more "small batches" than AMDs.

Oxide Games has emailed us this evening with a bit more detail about what's going on under the hood, and why Mantle batch submission times are higher. When working with large numbers of very small batches, Star Swarm is capable of throwing enough work at the GPU such that the GPU's command processor becomes the bottleneck. For this reason the Mantle path includes an optimization routine for small batches (OptimizeSmallBatch=1), which trades GPU power for CPU power, doing a second pass on the batches in the CPU to combine some of them before submitting them to the GPU. This bypasses the command processor bottleneck, but it increases the amount of work the CPU needs to do (though note that in AMD's case, it's still several times faster than DX11).

http://www.anandtech.com/show/8962/the-directx-12-performance-preview-amd-nvidia-star-swarm/6

 

[Stuka87]

I still think that Mantle was basically DX12 alpha. Both AMD and nVidia were involved with its development, and I think mantle was basically the proof of concept.

 

[DiogoDX]

So DX12 and GlNext = Mantle 1.0 features. In this case SDK makes no sense and the developer is better working on DX12 and GlNext.


Means they are working on 2.0?

 

[Noctifer616]

So DX12 and GlNext = Mantle 1.0 features. In this case SDK makes no sense and the developer is better working on DX12 and GlNext.


Means they are working on 2.0?

More info on Thursday.

 

[ShintaiDK]

So DX12 and GlNext = Mantle 1.0 features. In this case SDK makes no sense and the developer is better working on DX12 and GlNext.

What AMD seems to says is that Mantle will never be finished with the tools needed and developers should look for other APIs that got this. (SDK)

Means they are working on 2.0?

By the looks of it all development is stopped and all focus is on DX12.

 

[parvadomus]

Nope, just better hardware with DX12.
nVidia's command processor can process much more "small batches" than AMDs.


http://www.anandtech.com/show/8962/the-directx-12-performance-preview-amd-nvidia-star-swarm/6

That is pure speculation and a lot of nonsense. When they apply that software trick the fps just goes up only a little. Its rather obvious the gap between the 290 and 980 its not only related to that, it can be anything from drivers to any part of the hardware pipeline like pixelrate.
BTW this is the best that could happen to AMD for both their GPUs and CPUs.

 

[SlowSpyder]

I think the writing has been on the wall for a while. DX12 brings some of the same benefits. Mantle just brought some of those benefits early to GCN card owners in non-DX12 games.

 

[NTMBK]

If DX12 delivers the same benefits, good.

 

[Silverforce11]

AMD said in the past they were working to improve Mantle (2.0) with asynchronous DMA & compute which reduces GPU bottlenecks, but that never came to effect, and now we know these two features are going to be in DX12... guess that is where their effort needs to focus on.

Its not wise to split driver teams across too many APIs. Even if DX12 is widespread later, DX11 will still remain the bulk and it needs work. Mantle has done its job & should definitely be retired, its the wise choice.

 

[Phynaz]

Everyone who didn't see this coming raise your hand.

 

[Phynaz]

AMD said in the past they were working to improve Mantle (2.0) with asynchronous DMA & compute which reduces GPU bottlenecks, but that never came to effect, and now we know these two features are going to be in DX12... guess that is where their effort needs to focus on.

Its not wise to split driver teams across too many APIs. Even if DX12 is widespread later, DX11 will still remain the bulk and it needs work. Mantle has done its job & should definitely be retired, its the wise choice.

How do you retire a product that never made it out of beta? Cancelled is the more accurate term.

 

[Noctifer616]

How do you retire a product that never made it out of beta? Cancelled is the more accurate term.

We actually don't know that. For all we know Mantle 1.0 could have been finished months ago, but with it being under NDA there is just no way of knowing that.

I guess we will know more later this week.

 

[zlatan]

Don't worry guys. Mantle 1.0 just get a new name and Khronos will release the specs.