AMD GPU14 Tech Event Sept 25 - AMD Hawiian Islands

[Jaydip]

I already predicted AMD needs exclusives and here you are.Where are you Finalty?
I won't be surprised if NV comes with their own "Dismantle" stuff pretty soon

 

[wand3r3r]

Like what? Care to point it out and correct it? It's not that bad is it...

Well I can't read the future so there's nothing to prove.

Just a lot of doom and gloom from the usual suspect(s), complaining about "openness" when it fits (ignoring it when it's the other side etc.) their agenda. Lots of hate on the new mantle framework. Complaining about everything presented. Goalposts being shifted etc. Half of the people in here seem to have read a marketing blurb and since it's the "wrong" team (or something?) they declare it an instant failure.

I think that if "mantle" is a tremendous success, NV will pick up on it and suddenly the loudest voices in this thread will agree it's pretty neat. Not that I expect that, just pointing out the FUD in this thread.

I won't make an opinion until I have a card in my system with the new features so I can see if it actually works and how it is implemented. I actually think a lot of this is somewhat interesting, but will wait to see how it plays out.

It's hard to make an opinion on glorified marketing slides, but it appears that there will be some interesting developments in the near future. Whether they pan out remains to be seen.

 

[boxleitnerb]

This seems to be like PhysX, just way worse. Although the potential rewards are also way greater. Difficult to judge, we need more info. But the Mantle initiative with a combined Intel/AMD/NV effort would have been preferred.

 

[wand3r3r]

<snip />
But the Mantle initiative with a combined Intel/AMD/NV effort would have been preferred.

The same could be said of physx (and any other single manufacturer initiative), right? () At least, to my understanding, it's open for NV to jump on board. Nothing is stopping them, I haven't seen any sign that they have to pay for anything to AMD (unlike the reverse).

 

[SiliconWars]

This seems to be like PhysX, just way worse. Although the potential rewards are also way greater. Difficult to judge, we need more info. But the Mantle initiative with a combined Intel/AMD/NV effort would have been preferred.

Way worse how....?

 

[Fire&Blood]

As users, what do we have to lose? If Mantle doesn't gain traction, it's business as usual. And if it does have an impact, then Matle will address a problem that Nvidia has had to deal with as well and they will jump on board too.

 

[sontin]

As users, what do we have to lose? If Mantle doesn't gain traction, it's business as usual. And if it does have an impact, then Matle will address a problem that Nvidia has had to deal with as well and they will jump on board too.

So you want to go back in the mid 90s where we had different APIs for different hardware vendors and you need to check if a game support your hardware?

I guess a lot of people are still to young or were not interested in PC gaming at this time.

 

[SiliconWars]

So you want to go back in the mid 90s where we had different APIs for different hardware vendors and you need to check if a game support your hardware?

I guess a lot of people are still to young or were not interested in PC gaming at this time.

There will always be the DX option. It'll just be the slow, Nvidia option.

 

[boxleitnerb]

Way worse how....?

Proprietary. Yes, technically it is "open", but it still is focused on GCN. And GCN is AMDs architecture, thus proprietary.
Let's assume for a moment that Nvidia had offered PhysX to AMD (we'll never know for sure if they did or didn't, but let's assume). Optimized for CUDA and CUDA is Nvidias API, thus proprietary.

The difference is, that Mantle likely involves optimizations on a much wider scale and not just targets a couple of effects in 1-2 games/year.

The same could be said of physx (and any other single manufacturer initiative), right? () At least, to my understanding, it's open for NV to jump on board. Nothing is stopping them, I haven't seen any sign that they have to pay for anything to AMD (unlike the reverse).

Yes, but the impact of PhysX is much much smaller than the impact of Mantle would be. This could split the market in half much more easily than a couple of PhysX games ever could.

There will always be the DX option. It'll just be the slow, Nvidia option.

There will always be the CPU PhysX option. It'll just be the slow, AMD option

 

[SiliconWars]

Proprietary. Yes, technically it is "open", but it still is focused on GCN. And GCN is AMDs architecture, thus proprietary.
Let's assume for a moment that Nvidia had offered PhysX to AMD (we'll never know for sure if they did or didn't, but let's assume). Optimized for CUDA and CUDA is Nvidias API, thus proprietary.

The difference is, that Mantle likely involves optimizations on a much wider scale and not just targets a couple of effects in 1-2 games/year.

I'm not convinced that an API that makes games run faster with no obvious drawbacks, and is open enough that Nvidia simply needs to change their architecture to suit, is "way worse" than PhysX or even a bad thing at all.

Ask yourself this - would it be bad if Nvidia didn't exist? Clearly not. Therefore you are simply hating on it because it can be used by AMD to gain performance leverage over Nvidia. That's what the name of the game is.

 

[Fire&Blood]

So you want to go back in the mid 90s where we had different APIs for different hardware vendors and you need to check if a game support your hardware?

I guess a lot of people are still to young or were not interested in PC gaming at this time.

I turned 37 past June so I am well aware of your concerns here. I am more focused on actual benefits, if Mantle takes off then it's legit and beneficial to all of us. Even if things get to be as bad as you assume, nowadays checking whether a game supports my hardware will take just as much effort as making sure I meet the min. requirements in the non Mantle world.

 

[DarkKnightDude]

http://www.maximumpc.com/amd_r9_290x_will_be_much_faster_titan_battlefield_4

We had a chance to sit down with AMD Product Manager Devon Nekechuck to see how AMD&#8217;s new top dog R9 290X stacks up against the green team&#8217;s best single-GPU offering. According to Nekechuck, even though the R9 290X uses a 438 square mm die, which is significantly smaller than the Titan&#8217;s 550 sq. mm GK110 offering, it &#8220;will definitely compete with the GTX 780 and Titan.&#8221; When we asked what this means in real-world terms, he stated, &#8220;with Battlefield 4 running with Mantel (AMD&#8217;s new graphics API), the card will be able to &#8216;ridicule&#8217; the Titan in terms of performance.&#8221;

I'll be honestly astonished if it did.

 

[boxleitnerb]

I'm not convinced that an API that makes games run faster with no obvious drawbacks, and is open enough that Nvidia simply needs to change their architecture to suit, is "way worse" than PhysX or even a bad thing at all.

Ask yourself this - would it be bad if Nvidia didn't exist? Clearly not. Therefore you are simply hating on it because it can be used by AMD to gain performance leverage over Nvidia. That's what the name of the game is.

You could say the exact same thing about PhysX. AMD could have changed their architecture to support CUDA and PhysX.

Clearly it would be. Because then we would have a monopoly and you don't want that. Don't turn this around on me and get personal but look for the hidden dangers beyond the marketing speech.

Don't get me wrong - the low level API itself is a welcome development, but the way it is presented and made is no better than PhysX/Glide.

 

[sontin]

I turned 37 past June so I am well aware of your concerns here. I am more focused on actual benefits, if Mantle takes off then it's legit and beneficial to all of us. Even if things get to be as bad as you assume, nowadays checking whether a game supports my hardware will take just as much effort as making sure I meet the min. requirements in the non Mantle world.

I think you dont know how "great" it was in the 90s: Metal, Glide, OpenGL with certain vendor specific extension etc. You cant play these games. You dont need to wait for a driver update because the developer need to rewrite the renderer.

Every close API is a sign of desperation because your hardware cant compete on a even field. 3DFX showed it in perfection what happen if you dont need better hardware because competition cant use the own render path.

 

[Carfax83]

I think you guys are making a mistake by comparing PhysX to Mantle.

PhysX is just middleware that runs on CUDA. Mantle is an architecture specific API, so there's no way it can be classified as "open.".

You could port PhysX over to OpenCL and AMD would be able to run it without needing to use CUDA.

But how can NVidia use Mantle when it's created exclusively for GCN? If anything, Mantle is more like CUDA.

I don't see any way for NVidia to jump on the Mantle bandwagon, without making their own similar exclusive API that allows for deeper access on their hardware.

And so we'll end up with multiple APIs and some games running inherently much faster on NVidia hardware, and some running inherently much faster on AMD hardware.

Although I recognize the desire for developers to circumvent the software limitations that APIs like OpenGL and DirectX impose on hardware, AMD is really opening up a can of worms with this.

If it causes Intel and NVidia to retaliate, then what?

 

[SiliconWars]

Let's be brutally honest about it - Nvidia would do the same thing (and they probably already have their own in development as we speak). The only "issue" here is that AMD did it first.

 

[Silverforce11]

http://www.maximumpc.com/amd_r9_290x_will_be_much_faster_titan_battlefield_4

I'll be honestly astonished if it did.

Eh, if its anything like advertised, Mantle through a similar grunt hardware should yield massive gains compared to DX, so thats not surprising at all, its the way coding to the metal is supposed to work.

Without Mantle, if it matches Titan, then AMD did very well for a refresh. Given die space and TDP constraints if they can achieve 7970 Ghz + 30%, then its mission accomplished.

 

[Silverforce11]

If it causes Intel and NVidia to retaliate, then what?

Which game developers would Intel convince to code specifically for their HD series of iGPUs?

Likewise, NV can only convince TWIMTBP devs to throw in superficial PhysX thus far with very few good implementation.

The point is the major console wins for GCN. This means major developers will already be coding for it, and if Mantle eases or improve their efficiency for the PC market as a side-effect, very few devs would say no to that.

 

[boxleitnerb]

Let's be brutally honest about it - Nvidia would do the same thing (and they probably already have their own in development as we speak). The only "issue" here is that AMD did it first.

Such a cheap speculative shot. If Nvidia does it it's bad. If AMD does it, it's not bad. With that opinion I cannot take you seriously anymore.

Which game developers would Intel convince to code specifically for their HD series of iGPUs?

Rome 2, Grid 2 (Iris Pro exclusive features).

 

[Fire&Blood]

Every close API is a sign of desperation because your hardware cant compete on a even field. 3DFX showed it in perfection what happen if you dont need better hardware because competition cant use the own render path.

Although I recognize the desire for developers to circumvent the software limitations that APIs like OpenGL and DirectX impose on hardware, AMD is really opening up a can of worms with this.

If Mantle is a joke, it will be exposed before Q1 and it's back to business as usual. Assuming it's not and Nvidia can't/won't jump on board then AMD pulled off a coupe by securing the business of the money makers and pushing GCN/Mantle into both next gen consoles. I'm still not convinced it's as grim as it looks. I see it as less of a anti Nvidia move and more of a push against Microsoft that everyone seems to be doing these days.

 

[sontin]

AMD is doing nothing here. They are promoting an own API. There is no desire for a lot of developers to want this. Mantle was specific designed with Dice. Because these guys wanted something like this. You know the Johnny guy said he loves PC gaming and yet he is introducing a new API with will seperate the PC market.

Hyprocrite at best.

 

[Rikard]

So in summary:

• There are some new cards coming. They are not very hot tbh. Better just get current gen cards imo.
• Hardware enabled positional audio with direct access to game information looks great and should be much better than the 3rd party software we have now. Finally there is something happening in that area! Too bad it will only be on two new cards though. Combined with the *meh* of the above bullet, I personally will wait at least one more generation before jumping on that train though.
• Mantle looks promising, and I might get BF4 just to try it out on my existing GCN card. Initially it will be Windows only, but once it gets to Linux I might be able to retire Windows for good, depending on how it performs.

 

[itsmydamnation]

AMD is doing nothing here. They are promoting an own API. There is no desire for a lot of developers to want this. Mantle was specific designed with Dice. Because these guys wanted something like this. You know the Johnny guy said he loves PC gaming and yet he is introducing a new API with will seperate the PC market.

Hyprocrite at best.

http://www.bit-tech.net/hardware/graphics/2011/03/16/farewell-to-directx/1

'It's funny,' says AMD's worldwide developer relations manager of its GPU division, Richard Huddy. 'We often have at least ten times as much horsepower as an Xbox 360 or a PS3 in a high-end graphics card, yet it's very clear that the games don't look ten times as good. To a significant extent, that's because, one way or another, for good reasons and bad - mostly good, DirectX is getting in the way.' Huddy says that one of the most common requests he gets from game developers is: 'Make the API go away.'

'I certainly hear this in my conversations with games developers,' he says, 'and I guess it was actually the primary appeal of Larrabee to developers &#8211; not the hardware, which was hot and slow and unimpressive, but the software &#8211; being able to have total control over the machine, which is what the very best games developers want. By giving you access to the hardware at the very low level, you give games developers a chance to innovate, and that's going to put pressure on Microsoft &#8211; no doubt at all.'

http://forum.beyond3d.com/showpost.php?p=1788711&postcount=210

Mantle is the direct result of a number of AAA game developers asking us for something like this because of several constraints they're experiencing with existing PC graphics APIs. Those developers want to get closer to the metal and able to achieve more with the GPU(s) on their PC without the API getting in the way of this effort. "More with more" is the term used by Raja and it couldn't be more spot on.
AMD listened to developer feedback and entered this venture with Johan Andersson leading the charge. Johan has a passion for writing efficient code that can allow him to accomplish his technical vision and Mantle exists to serve this purpose.

Some technical concerns are mentioned on this thread. The Mantle API has been in design and development for more than 2 years and rest assured we've had some time to think about the best way to address those (and loads of others).

got any more lies to make up sontin

this is also a very good read
http://www.beyond3d.com/content/articles/120/1

 

[SiliconWars]

What I don't get is how come the "software company" that JHH claimed Nvidia was, didn't get there first?

You know the answer? Nvidia resting on laurels, trying to make Tegra relevant while AMD made themselves the real gaming company in the past two years.

 

[sontin]

http://www.bit-tech.net/hardware/graphics/2011/03/16/farewell-to-directx/1
http://forum.beyond3d.com/showpost.php?p=1788711&postcount=210

got any more lies to make up sontin

this is also a very good read
http://www.beyond3d.com/content/articles/120/1

So, an AMD site (beyond3d) and an AMD employee are promoting their own vendor specific API.

Yap, i guess i should take this serious. :\

What I don't get is how come the "software company" that JHH claimed Nvidia was, didn't get there first?

You know the answer? Nvidia resting on laurels, trying to make Tegra relevant while AMD made themselves the real gaming company in the past two years.

Lol, the reason why nVidia did not do it was it will seperate the pc market and will not get support outside of partner they are paying. Nobody will use so many different APIs. The reason why Dice is using it is because they wanted it. It's a console thing for a closed plattform.