What exactly does AMD get out of continued mantle development post dx12?

T

tybert7

2[H]4U
Joined
Aug 23, 2007
Messages
2,763
So many of us have probably seen the two Huddy interviews already. In each I believe he mentioned that while amd would open mantle up to both nvidia and intel, the control over the direction of mantle would remain with AMD. On top of that, several devs wanted AMD to continue supporting mantle even after dx12 launched.


My question is, why the last?

I get why amd wants to stay in control of the mantle spec, it can probably be tuned to scale best on AMD engineered hardware giving them a technical edge anytime that api is used. That is also a reason beyond pride that a company like nvidia might not want to tie its fate to an api where the hardware perks are best gained with amd cards.

But what kind of extra advantages will mantle bring over dx12 after it launches? Will it still be even more efficient at running games? Will it be as Huddy suggested, a way to iterate on features faster than the dx12 spec would allow?

If yes to the last, what kind of features? Real time diffraction? ray tracing?

Is mantle going to be the personal test bed of Johan for exotic new features? And if so why could that not be done as well with dx12?

I'm not complaining, I'm in the amd camp so this can only help me, but I'm just curious about why.
 
DX12 will be at minimum Windows 8.1, so it will have a small market share.

DX11 has yet to be adopted properly, nobody is using the DX11 features, so what it the point of DX12.
 
Spazturtle said:
DX12 will be at minimum Windows 8.1, so it will have a small market share.

DX11 has yet to be adopted properly, nobody is using the DX11 features, so what it the point of DX12.
Click to expand...

Its got to exist before it can be used, even if it a slow uptake.
 
Mantle could be an easy way to port games to Linux and potentially SteamOS.

Quite a Big opportunity right there!
 
Yakk said:
Mantle could be an easy way to port games to Linux and potentially SteamOS.

Quite a Big opportunity right there!
Click to expand...

There are no games written only in Mantle. So the game would have to first be ported from DX to Mantle and then ported again to Linux. Once on Linux it would only run on a small minority of systems as Intel and NVIDIA dominate the market.

It would make more sense to either write it in OpenGL to begin with or port it to OpenGL which has universal support on Linux.

Assuming we even see Mantle for Linux. DX12 and OpenGL 4.4 will be the chosen API for newer game engines. Mantle will be a patch as long as AMD continues to fund its inclusion in games.
 
hogwash.

developers will port thier games to mantle to Linux without a second thought.

Most of the major engines (except UE - Sweeny is a tool anyway) already support mantle natively, no patch required. And as Theif has shown, patching it is pretty easy.

40% of the Decrete GPU linux the market is better than 0%
 
Final8ty said:
Adding Mantle API and True Audio to game that was not planned with Mantle API and True Audio at conception that use an engine with no native Mantle support in 2 and a half weeks after release is impressive.
Click to expand...

I'm not sure why hes still getting quoted, it hurts my brain to read his texts in quotes.

Also Mantle is an effort to get better developer relations. Mantle doesn't have a single proponent that will help AMD out specifically and not its competition (well right now it does with GPU's).

Developers were asking for this for a while so they gave them something they wanted, its very well known that AMD has had issues with maintaining developer relations and I believe Mantle is a good way to get their foot back into that door. Especially considering developers were begging for something like this for quite some time now.
 
Trimlock said:
I'm not sure why hes still getting quoted, it hurts my brain to read his texts in quotes.

Also Mantle is an effort to get better developer relations. Mantle doesn't have a single proponent that will help AMD out specifically and not its competition (well right now it does with GPU's).

Developers were asking for this for a while so they gave them something they wanted, its very well known that AMD has had issues with maintaining developer relations and I believe Mantle is a good way to get their foot back into that door. Especially considering developers were begging for something like this for quite some time now.
Click to expand...

Indeed.
 
Mantle is purely the first API to compete with DirectX I'm over a decade. I'm excited at the prospect of Microsoft actually having competition.

OpenGL does not count.
 
KazeoHin said:
Mantle is purely the first API to compete with DirectX I'm over a decade. I'm excited at the prospect of Microsoft actually having competition.

OpenGL does not count.
Click to expand...

If Dx12 is even equal to Mantle in its present state i will be happy, but im hoping that it will be better.
Oh wait, is Dx12 coming to windows 7 or is that still a no go ?
 
Final8ty said:
If Dx12 is even equal to Mantle in its present state i will be happy, but im hoping that it will be better.
Oh wait, is Dx12 coming to windows 7 or is that still a no go ?
Click to expand...

The official word is that DX12 will be able to run on the same sort of hardware that Windows 7 can run on. Whatever that means.
 
King Icewind said:
Increased performance on their APUs (CPU/GPU).
Click to expand...

I see this for showing off Mantle's abilities best. It is on APU powered devices/computers that Mantle would benefit the most, leveraging workload off the CPU better for these type of devices. Improved gaming experience on ultra books and low-end laptops and such without discrete graphics.
 
KazeoHin said:
Mantle is purely the first API to compete with DirectX I'm over a decade. I'm excited at the prospect of Microsoft actually having competition.

OpenGL does not count.
Click to expand...

I personally believe that without Mantle as competition, we would not be seeing DX12 now.
 
Brent_Justice said:
I personally believe that without Mantle as competition, we would not be seeing DX12 now.
Click to expand...

I agree wholeheartedly. Less than two years ago, AMD said that they had no knowledge of the next 'iteration' of DirectX(12) and that they had not been approached by Microsoft to discuss the idea.

Now Microsoft claims that DX12 has been 'in development' for 'over four years' ?
 
tybert7 said:
My question is, why the last?
Click to expand...

Because the high end game market survives by being on the leading edge of technology. They differentiate themselves from the competition by being either the first to offer support for new hardware features or making a better performing implementation of them in case of feature parity with competing games.

Mantle allows AMD to expose any new hardware feature on upcoming hardware through the API to those developers without having to wait for Microsoft to get off their ass and add it to DirectX. As Huddy explained, there can be games being developed with support for brand new hardware features before that actual hardware is released. That is good for both the developer and the hardware vendor.

P.S: And this is not limited to AMD, Intel and Nvidia will benefit since they can just update their Mantle driver and expose any new hardware feature as well.
 
I see a few key things Mantle will do for AMD:

1. Short-term, improve performance of its APUs compared to Intel. It always seems like each generation of AMD's APUs are on the cusp of providing an all-in-one solution for consumers that would take over the low-end market (tablets, school laptops, netbooks, etc.). Maybe this will provide enough of a push to edge out Intel a little and give AMD some breathing room.

2. Mid-term, it will help them gain an edge of nVidia. My slightly biased opinion is that AMD provides better hardware, but nVidia works better with game devs for "optimizations", so AMD never quite gains as much market share as it should. Driver history also plays into this of course. S even if it is an open standard and everyone contributes, I still think this would benefit AMD hardware more than nVidia's.

3. Long-term, hopefully it will enable everyone the option of switching to Linux, breaking Microsoft's monopoly with DirectX. Although the gaming situation keeps slowly improving on Linux, this could potentially provide a better experience than Windows, which is what I predict it'll take to get gamers to switch over. I think that's a pretty uncontroversial opinion, and it's why Wine hasn't really taken off as much as I'd like. Also, long-term they will share development costs with others, such as nVidia, Intel, and game devs. Hopefully they're honest about this, because otherwise I see it failing. At this point though, they will have hopefully gained a large chunk of market share and consumer goodwill.

Once that happens, Microsoft either steps up its game or DX dies and this "open" API takes over. We're already seeing the first stages of this with the SteamBox, where another company is confident enough to bring a product forward.
 
Andrew_Carr said:
I see a few key things Mantle will do for AMD:

1. Short-term, improve performance of its APUs compared to Intel. It always seems like each generation of AMD's APUs are on the cusp of providing an all-in-one solution for consumers that would take over the low-end market (tablets, school laptops, netbooks, etc.). Maybe this will provide enough of a push to edge out Intel a little and give AMD some breathing room.

2. Mid-term, it will help them gain an edge of nVidia. My slightly biased opinion is that AMD provides better hardware, but nVidia works better with game devs for "optimizations", so AMD never quite gains as much market share as it should. Driver history also plays into this of course. S even if it is an open standard and everyone contributes, I still think this would benefit AMD hardware more than nVidia's.

3. Long-term, hopefully it will enable everyone the option of switching to Linux, breaking Microsoft's monopoly with DirectX. Although the gaming situation keeps slowly improving on Linux, this could potentially provide a better experience than Windows, which is what I predict it'll take to get gamers to switch over. I think that's a pretty uncontroversial opinion, and it's why Wine hasn't really taken off as much as I'd like. Also, long-term they will share development costs with others, such as nVidia, Intel, and game devs. Hopefully they're honest about this, because otherwise I see it failing. At this point though, they will have hopefully gained a large chunk of market share and consumer goodwill.

Once that happens, Microsoft either steps up its game or DX dies and this "open" API takes over. We're already seeing the first stages of this with the SteamBox, where another company is confident enough to bring a product forward.
Click to expand...


When AMD said they would open up Mantle for a public SDK (if nothing happens that would interfere) this year I believe them.
The hardware platforms where AMD is under performing will change as long as there is a Mantle version.

Doubt that the Nvidia problem is really a problem when they can use Mantle later on this year.

The best thing about Mantle is that when hardware becomes exotic (multiple gpu) it is still your best bet to get good if not outstanding performance from it. This is one of the main reasons why Mantle on "all" platforms still has a good change of being there even when DX20 hits ;) .
 
Last edited:
You must log in or register to reply here.
Back
Top