The AMD Mantle Thread

[bystander36]

Do you think they haven't considered everything you've stated here?

I don't think that many have looked beyond AMD approaching them, offering them some extra money to help get it started, and the temptation of being able to code on a PC closer to the metal than before. And for that, it is worth it to them.

I'm thinking more long term. You have to realize that if AMD does it, that forces Nvidia to do it. Now things get more convoluted.

Down the road, architectures change. New API's have to be made, or existing ones must change. Now it gets more convoluted, more fragmented.

Mantle seems short sighted to me. It may seem great now, but it is opening a can of worms. One we'll be dealing with in a couple years.

 

[SiliconWars]

Are you messing Mantle and consoles together now again?

It's called GCN. You know the architecture that Mantle and consoles both work on.

/topic
Funny how this "easy to implement with low man hours"...seems to drag out another month...color me surprised

If you read the press release you'd know that it's basically ready. They decided to delay the launch, they didn't have to delay it.

Don't worry, you'll see plenty of examples at CES in just over a week.

 

[krumme]

You can't seriously believe you own reasoning in that last paragraph can you? Mantle will be limited to perhaps 10 to 15 percent of the PC market, while DX is applicable to all PCs not running IoS or Linux, which must be well above 90% of the market. To say they have similar compatibility issues is simply absurd.

How many plays bf4 on an Intel gpu?

Mantle plus consoles is easily 85% of the total market for bf4 in the next 2 years. There will be sold M of kaveris.

If EA is something its greedy. They are only going full steam into Mantle because of selling more games and more expensive games. Its very simple.

Its aparent from Johans speak at Apu13 that the 15 games is only whats is disclosed for now. More is comming.

Secondly its was very aparent from his plants vs zombies that lower end apu like kaveri is the target for some of the less demanding although kaveri also "plays bf4 really well".

Perhaps EA knows their market. At least they are using mantle to get an compettitive edge.

 

[SiliconWars]

Down the road, architectures change. New API's have to be made, or existing ones much change. Now it gets more convoluted, more fragmented.

Mantle seems short sighted to me. It may seem great now, but it is opening a can of worms. One we'll be dealing with in a couple years.

This has been covered. It's just the right level of abstraction so that it can be adapted to future architectures "and possibly even our competitors hardware"

http://www.youtube.com/watch?feature=player_detailpage&v=tDPgJB2x7dQ#t=127

 

[3DVagabond]

I don't think that many have looked beyond AMD approaching them, offering them some extra money to help get it started, and the temptation of being able to code on a PC closer to the metal than before. And for that, it is worth it to them.

I'm thinking more long term. You have to realize that if AMD does it, that forces Nvidia to do it. Now things get more convoluted.

Down the road, architectures change. New API's have to be made, or existing ones must change. Now it gets more convoluted, more fragmented.

Mantle seems short sighted to me. It may seem great now, but it is opening a can of worms. One we'll be dealing with in a couple years.

You are just repeating your concerns here. I understand what you said. They most assuredly would have considered all of this.

 

[Lonbjerg]

It's called GCN. You know the architecture that Mantle and consoles both work on.

There is NO Mantle in ANY console...stop that debunked lie thanks.

 

[Lonbjerg]

How many plays bf4 on an Intel gpu?

Mantle plus consoles is easily 85% of the total market for bf4 in the next 2 years. There will be sold M of kaveris.

If EA is something its greedy. They are only going full steam into Mantle because of selling more games and more expensive games. Its very simple.

Its aparent from Johans speak at Apu13 that the 15 games is only whats is disclosed for now. More is comming.

Secondly its was very aparent from his plants vs zombies that lower end apu like kaveri is the target for some of the less demanding although kaveri also "plays bf4 really well".

Perhaps EA knows their market. At least they are using mantle to get an compettitive edge.

They also said Mantle would come december...kinda not trusting Mantle PR...based on it's failure to deliver.
Empty words...don't run a game.

 

[SiliconWars]

How many plays bf4 on an Intel gpu?

Mantle plus consoles is easily 85% of the total market for bf4 in the next 2 years. There will be sold M of kaveris.

If EA is something its greedy. They are only going full steam into Mantle because of selling more games and more expensive games. Its very simple.

Its aparent from Johans speak at Apu13 that the 15 games is only whats is disclosed for now. More is comming.

Secondly its was very aparent from his plants vs zombies that lower end apu like kaveri is the target for some of the less demanding although kaveri also "plays bf4 really well".

Perhaps EA knows their market. At least they are using mantle to get an compettitive edge.

Not only that, but it's clear that DICE at least are putting DX firmly in the 2nd class seat.

http://image.slidesharecdn.com/joha...47-phpapp01/95/slide-31-638.jpg?cb=1384519187

"Mantle and PlayStation 4 will drive our future Frostbite designs & optimizations"

It's just too easy to get a PC game from console to Mantle compared with console to DX. This is AMD's big advantage and I'm not sure if people truly realise just how game changing this aspect is.

AMD is basically in the position to pay EA to release PC games first on GCN. Will they do it? Yes I believe they will. That's when the real fun starts.

 

[bystander36]

This has been covered. It's just the right level of abstraction so that it can be adapted to future architectures "and possibly even our competitors hardware"

http://www.youtube.com/watch?feature=player_detailpage&v=tDPgJB2x7dQ#t=127

Then that limits AMD's possible architecture changes. "and possibly even our competitors hardware" would suggest it might not. It also means it would limit their hardware architecture changes.

Anyways, you get my point. This is the start of fragmentation on the PC. I don't think this a good thing. There are many dev's that feel the same.

We'll have to wait and see how it plays out, but don't be surprised if things don't turn out well. Fewer PC ports/games, buggier PC games and higher dev costs are all very possible issues we'll face.

 

[bystander36]

You are just repeating your concerns here. I understand what you said. They most assuredly would have considered all of this.

Why would they care. They got their pay day. They got their chance to code closer to the metal. I doubt they care about problems in the future.

 

[SiliconWars]

There is NO Mantle in ANY console...stop that debunked lie thanks.

Do you have any idea of the concepts in this discussion?

GCN = Graphics Core Next. You know the actual graphics hardware used in the consoles and AMD graphics cards + APU's?

They are the same.

The API in the PS4 is extremely similar to Mantle. No it's not "identical" but it's so similar that porting between each is probably a matter of DAYS. The XBox is likely not much different.

It doesn't need to be identical in order to be much easier. This is not difficult to understand Lonbjerg so try harder.

 

[krumme]

Who says nv is interested in making a new api for x86?

Igpu is slowly killing them in x86 market especially for the notebooks. There is a reason they invest heavily in tegra arm business.

For amd it also looks like they are going away from x86 market albeit much slower. Mantle looks like a part of the platform for hsa and then their first entry to arm world.

 

[SiliconWars]

Then that limits AMD's possible architecture changes. "and possibly even our competitors hardware" would suggest it might not. It also means it would limit their hardware architecture changes.

Yes as I said, this is the decision that AMD made so that Mantle could work. It is a sacrifice of sorts but they clearly believe it is worth it.

Anyways, you get my point. This is the start of fragmentation on the PC. I don't think this a good thing. There are many dev's that feel the same.

I just see it as more performance for me. The devs will not write for Mantle out of the goodness of their hearts, they will only do it if there is an incentive.

We'll have to wait and see how it plays out, but don't be surprised if things don't turn out well. Fewer PC ports/games, buggier PC games and higher dev costs are all very possible issues we'll face.

Again, you're only seeing it from the Nvidia side. This is not going to affect me. That might sound callous but for the past few years I've had nothing but BS in some games because of Nvidia and their partners hobbling AMD cards on purpose.

I could have bought an Nvidia card back then and not suffered it. You and everybody else can buy an AMD card and not suffer the fallout from Mantle. It's your choice but I feel no sympathy for anyone on the green side same as I got none.

 

[bystander36]

Who says nv is interested in making a new api for x86?

Igpu is slowly killing them in x86 market especially for the notebooks. There is a reason they invest heavily in tegra arm business.

For amd it also looks like they are going away from x86 market albeit much slower. Mantle looks like a part of the platform for hsa and then their first entry to arm world.

If Mantle is what you guys seem to think it'll be, they'll have no choice. Of course the jury is out on it.

 

[bystander36]

Again, you're only seeing it from the Nvidia side. This is not going to affect me. That might sound callous but for the past few years I've had nothing but BS in some games because of Nvidia and their partners hobbling AMD cards on purpose.

I could have bought an Nvidia card back then and not suffered it. You and everybody else can buy an AMD card and not suffer the fallout from Mantle. It's your choice but I feel no sympathy for anyone on the green side same as I got none.

How is this only an Nvidia problem? If we have several different API's, the problems will be shared across the board. Nvidia sponsored games will likely get preferential treatment, the same for AMD games. The whole platform becomes less attractive to dev's, as they have to spread their focus across 3 API's.

 

[krumme]

How do people actually think a huge global company like EA comes to decisions to invest in a new engine that is mantle compatible. Its a damn serious investment. Covering bilions in investment forward. Its damn well thought out and analyzed. They know to the detail how much code is directly compatible and what is only similar. And thats what it is, because the consoles constitutes the huge part of the revenue. Its more or less the same. Otherwise we wouldnt see all those games comming. Its a freaking fast ramp of mantle and it can only be because is a minor investment.

 

[bystander36]

How do people actually think a huge global company like EA comes to decisions to invest in a new engine that is mantle compatible. Its a damn serious investment. Covering bilions in investment forward. Its damn well thought out and analyzed. They know to the detail how much code is directly compatible and what is only similar. And thats what it is, because the consoles constitutes the huge part of the revenue. Its more or less the same. Otherwise we wouldnt see all those games comming. Its a freaking fast ramp of mantle and it can only be because is a minor investment.

That doesn't change that it is opening up a bad precedent for future engines. It works for this engine. It opens the doors for more API's to do the same in the future.

I get that you are excited. You have a few hundred posts about it. I'm not excited about the consequences.

 

[SiliconWars]

How is this only an Nvidia problem? If we have several different API's, the problems will be shared across the board. Nvidia sponsored games will likely get preferential treatment, the same for AMD games.

This isn't much different from what we already have. From an AMD point of view it's not a problem because when the devs really start using the new consoles to their maximum, they will automatically be squeezing the most out of AMD graphics with Mantle as well.

I know what you're trying to say but you have to see it from the AMD-centric viewpoint. AMD simply does not care, AMD owners just don't care. We can only see upsides.

The whole platform becomes less attractive to dev's, as they have to spread their focus across 3 API's.

It's less attractive because of DX though - not Mantle. It's DX that is the problem here - this is why they are all saying that they really want to see Mantle on Nvidia hardware, so they can drop the money and time sink that is DX.

 

[Gloomy]

It's less attractive because of DX though - not Mantle. It's DX that is the problem here - this is why they are all saying that they really want to see Mantle on Nvidia hardware, so they can drop the money and time sink that is DX.

that statement is going to set some butts on this forum aflame. You may want to run for the hills before the MSDF finds you

 

[bystander36]

that statement is going to set some butts on this forum aflame. You may want to run for the hills before the MSDF finds you

His comment makes no sense, unless he think Mantle makes the PC platform more attractive. You still have to program for DX regardless of what happens.

 

[SiliconWars]

It's true though. Both Repi and the Oxide guys are basically desperate for Nvidia to adopt Mantle and it's because that would make their lives incredibly easy. That's what they are interested in - they no longer want to go through the extreme hassles of porting to DX after seeing what Mantle can do.

 

[krumme]

If Mantle is what you guys seem to think it'll be, they'll have no choice. Of course the jury is out on it.

I dont think its impossible nv will make their own api. But very close to that

Who should adopt that api and program for it?

I can only see devs only making games for pc. Take eg oxide ???

And what about the timespan. Amd have been working on this for 4 years until we get the first game. I mean where is the market in say 2-3 years from now?

Its a mess. The consoles is the driving force. They could wait for next gen cosoles in 8 years (not ps4.1 1080p ultra but ps5). And use it as a jump to get the consoles on arm and thereby getting consumers away from Intel monopoly. It must be a very tempting choise. But arm is also red ocean. Dont know, its aparent Mantle is a bold and disruptive move.

 

[bystander36]

It's true though. Both Repi and the Oxide guys are basically desperate for Nvidia to adopt Mantle and it's because that would make their lives incredibly easy. That's what they are interested in - they no longer want to go through the extreme hassles of porting to DX after seeing what Mantle can do.

Unless DX can be replaced, they still have to go through all the hoops. Since Mantle is GCN locked, it doesn't change that. It only adds to the hoops.

 

[SiliconWars]

Unless DX can be replaced, they still have to go through all the hoops. Since Mantle is GCN locked, it doesn't change that. It only adds to the hoops.

Then DX will be replaced. Until that point the devs can go with Mantle on AMD and use Nvidia's GameWorks libraries in order to make it easier on DX. The combination of those two might even be less effort than coding for DX under both vendors.

Nvidia is in full damage limitation mode until they get their own API out, at which point DX goes the way of the dodo. There is just no saving it now. That's what I mean by the cat being out the bag. It's started, no point wishing it hadn't any longer.

 

[bystander36]

Then DX will be replaced. Until that point the devs can go with Mantle on AMD and use Nvidia's GameWorks libraries in order to make it easier on DX. The combination of those two might even be less effort than coding for DX under both vendors.

Nvidia is in full damage limitation mode until they get their own API out, at which point DX goes the way of the dodo. There is just no saving it now. That's what I mean by the cat being out the bag. It's started, no point wishing it hadn't any longer.

No, DX still has to stay, unless Dev's abandon the AMD non GCN cards, and what ever Nvidia doesn't support with their API. There is still Intel as well.