AMD Mantle Graphics API Adopted by Various Developers

[ltguy]

The idea behind Mantle is that it becomes easier, not harder, to port games from one environment to another.

 

What if your pc doesnt have an ATI video card? Some people have these other graphics cards... NVIDIA   Is it still easier if you exist in a world where not all graphics cards are produced by AMD. If a game is going to be on pc, it must run on a wide range of hardware. Coding the game once for everyone and then once for mantle users isnt easier. Its wasted energy.

[Vitalius]

Win Win ?? This is an awful idea. Did you forget the 3dfx glide days ? They killed it for a reason. Physx is stupid for the same reason. I am totally against designing games to run on just one brand of video card. This makes about as much sense as making games exclusive to just one console. PC gaming is finally gaining some traction in the industry. That last thing we need to do is throw fragmentation into the mix. I will 100% boycott any game studio that supports mantle. If someone wants to design an open source replacement to DirectX that can work on either camps cards, then fine. But having to code games in mulitple ways for mutiple cards is a waste and is the last thing pc gaming needs right now. Not to mention the consoles dont support mantle so yet another rift in the mix. Consoles have finally aligned themselves with pcs so we can all play the same games and be fat dumb and happy and now amd wants to distance themselves from that so its even harder to code games once for pc/xboxone/ps4. This is DOA.

What if your pc doesnt have an ATI video card? Some people have these other graphics cards... NVIDIA   Is it still easier if you exist in a world where not all graphics cards are produced by AMD. If a game is going to be on pc, it must run on a wide range of hardware. Coding the game once for everyone and then once for mantle users isnt easier. Its wasted energy.

 

Ohhh, doom and gloom. I'm so scared.

You are talking about something you don't know anything about. Mind you, what I know is only from press releases and explanations on how it is designed to work. And that's all we can really go on.

Assuming it works as intended, Mantle will have built in ability to directly communicate with DirectX. Meaning games made in Mantle will work with Nvidia GPUs but without the performance benefit. Because Mantle is compatible with DirectX by design.

i would get an r9 290x if it was quiet but its not so i wont

 

... There is a quiet mode. And you can turn down the fan speed. You might sacrifice 3-5FPS by dropping it 5% but 35% fan speed is pretty much dead quiet.

There are non-reference models on the way.

Once again. Mantle is a PC implementation of the existing low-level APIs that are being used in the new generation consoles.

The idea behind Mantle is that it becomes easier, not harder, to port games from one environment to another.

This.

[Fetzie]

What if your pc doesnt have an ATI video card? Some people have these other graphics cards... NVIDIA   Is it still easier if you exist in a world where not all graphics cards are produced by AMD. If a game is going to be on pc, it must run on a wide range of hardware. Coding the game once for everyone and then once for mantle users isnt easier. Its wasted energy.

 

They already have to support OpenGL and DirectX. They're giving OpenGL access to some of the more pertinent features (for AMD cards ofc). They'll have to support Linux too when Steam's Linux client finally goes live. Why is nobody bitching about how Linux dev time is going to suck up resources like they are complaining about Mantle?

 

Why don't we wait for November 13 before speculating on how hard Mantle will be to implement (when we'll actually have facts on the table, and not just opinions)?

[ltguy]

They already have to support OpenGL and DirectX. They're giving OpenGL access to some of the more pertinent features (for AMD cards ofc). They'll have to support Linux too when Steam's Linux client finally goes live. Why is nobody bitching about how Linux dev time is going to suck up resources like they are complaining about Mantle?

 

Why don't we wait for November 13 before speculating on how hard Mantle will be to implement (when we'll actually have facts on the table, and not just opinions)?

I have complained about the linux dev time too. Steam OS is bad timing for pc gaming as well. More fragmentation. 

[ACatWithThumbs]

They already have to support OpenGL and DirectX. They're giving OpenGL access to some of the more pertinent features (for AMD cards ofc). They'll have to support Linux too when Steam's Linux client finally goes live. Why is nobody bitching about how Linux dev time is going to suck up resources like they are complaining about Mantle?

 

Why don't we wait for November 13 before speculating on how hard Mantle will be to implement (when we'll actually have facts on the table, and not just opinions)?

AMD has about 38% of the market share and more then a halph of those cards don't even support Mantle.

So Mantle doesn't make sense for anybody in the industry.

They need a diffrent low level solution like Nvidia with NVAPI wich is used for years now.

[Mo5]

While I'm excited, I'll start believing when I see the benchmarks  

[Vitalius]

AMD has about 38% of the market share and more then a halph of those cards don't even support Mantle.

So Mantle doesn't make sense for anybody in the industry.

They need a diffrent low level solution like Nvidia with NVAPI wich is used for years now.

 

As I told ltguy, according to AMD, Mantle is compatible with DirectX. I take this to mean that if a game were purely developed in Mantle, it would work with DirectX because Mantle knows how to communicate with it. Said game would not have the performance benefit of Mantle, but it would work with DirectX just fine.

While I'm excited, I'll start believing when I see the benchmarks  

Agreed.

[Fetzie]

AMD has about 38% of the market share and more then a halph of those cards don't even support Mantle.

So Mantle doesn't make sense for anybody in the industry.

They need a diffrent low level solution like Nvidia with NVAPI wich is used for years now.

Don't G-Sync and Shield require a GTX 650 or higher graphics card to work?

[kuddlesworth9419]

I wanna see actual figures though. If the performance is negligible then there is no point developing for it unless it saves time making the game. 

[bluedyguard]

AMD has about 38% of the market share and more then a halph of those cards don't even support Mantle.

So Mantle doesn't make sense for anybody in the industry.

They need a diffrent low level solution like Nvidia with NVAPI wich is used for years now.

Because nVidia TOOOOOOOOOOOOTALLY supports old cards. You have to keep your mind on the future, that's why AMD made Bulldozer.

[ShiverMePenguins]

Don't G-Sync and Shield require a GTX 650 or higher graphics card to work?

woah now, don't talk sense to the fanboys.

[digitalnav]

Support is one thing, we have yet to see any real world numbers. It's all talk at the moment guys.

[Gunjob]

AMD has about 38% of the market share and more then a halph of those cards don't even support Mantle.

So Mantle doesn't make sense for anybody in the industry.

They need a diffrent low level solution like Nvidia with NVAPI wich is used for years now.

NVAPI isn't the same as Mantle, NVAPI is more of a development tool that works with DX thus only small performance gains can be got, where as Mantle is a standalone low level API, which takes over the rendering as if it was OpenGL or DX. NVAPi isn't the same thing. 

[beebskadoo]

I hate to be a Debbie Downer but AMD is missing the damn point entirely!

Yeah you can brag all you want about your groundbreaking API that's going to revolutionize the industry by giving developers a console esque low level API for PC but at the end of the day this is all marketing BS.

Benchmarks AMD! 

Numbers AMD!

Proof!

 

That's how the world works, proof is needed or all of this is just marketing gobbledygook (never used that word in my life )

It's bad business to sell someone a product without letting them know how it will perform. A large part of their Hawaii reveal was focused on GCN Mantle support and their future strategy when it came to AMD GPU's+Mantle. 

All we want to know is what to expect when it comes to performance increases. That's all! Most of us don't want to hear marketing speak, we want numbers and proof!

[Immort]

Mantle is such a polarizing topic of discussion and because it's so polarizing fan-bias is highly prevalent, this creates an environment where misinformation spreads quickly and perception becomes fragile.
People who have been buying Nvidia products for the past... I don't know eight years, have made purchasing decisions that were largely influenced by the exclusive Nvidia features so for those people they are used to a brand that offered them most of the features they wanted that is until AMD announced Mantle, now there is a feature they want but can't have.

On the other hand, people who have been buying AMD products have been largely agnostic to features and believed performance/$ is king, to those people Mantle is the icing on the cake because it adds even more performance/$.

Your decisions have consequences, you can't rely anymore on one brand to give you everything you want, you actually have to make a real decision now, instead of paying a premium to forget about it.

[TechFan@ic]

Some people are shouting benchmarks, proof etc... AMD already announced, in the same article, that they're going to showcase performance in the upcoming event next week.

[beebskadoo]

Mantle is such a polarizing topic of discussion and because it's so polarizing fan-bias is highly prevalent, this creates an environment where misinformation spreads quickly and perception becomes fragile.

People who have been buying Nvidia products for the past... I don't know eight years, have made purchasing decisions that were largely influenced by the exclusive Nvidia features so for those people they are used to a brand that offered them most of the features they wanted that is until AMD announced Mantle, now there is a feature they want but can't have.

On the other hand, people who have been buying AMD products have been largely agnostic to features and believed performance/$ is king, to those people Mantle is the icing on the cake because it adds even more performance/$.

Your decisions have consequences, you can't rely anymore on one brand to give you everything you want, you actually have to make a real decision now, instead of paying a premium to forget about it.

You speak of 'misinformation'

That would be relevant if we actually had any real information on how Mantle performs. No one knows except AMD and the DEVS, and neither one of them are saying anything right now except for

'wow it's so cool and you're gonna love it! We promise'

bad business IMHO

[ACatWithThumbs]

As I told ltguy, according to AMD, Mantle is compatible with DirectX. I take this to mean that if a game were purely developed in Mantle, it would work with DirectX because Mantle knows how to communicate with it. Said game would not have the performance benefit of Mantle, but it would work with DirectX just fine.

Agreed.

That doesn't change anything you still have to make the game for Mantle to get a peformance gain wich no dev will do unless AMD gives them money to do so.

(Like Nvidia does with PhysX)

 

Don't G-Sync and Shield require a GTX 650 or higher graphics card to work?

Yes but those don't need Dev support

 

 

NVAPI isn't the same as Mantle, NVAPI is more of a development tool that works with DX thus only small performance gains can be got, where as Mantle is a standalone low level API, which takes over the rendering as if it was OpenGL or DX. NVAPi isn't the same thing. 

Yes but NVAPI gives a peformance gain for all Nvidia cards in every game wich Mantle won't.

I want to see the Dev that will support Mantle besides DX10/DX11/DX11.1/DX11.2/OpenGL/GCM/Low level Xbox DX11.2.

To think that any Dev will make Mantle support without getting buckets full of money from AMD is dream thinking.

Thats also why NVAPI is the better solution it gives a low level layer to DX instead of trying to replace it wich is in the PC market impossible.

Also the result of NVAPI is bigger then you think or how can it be that an TITAN is on par with a R9-290X wich has 1,1Tflop more compute power.

 

[beebskadoo]

Some people are shouting benchmarks, proof etc... AMD already announced, in the same article, that they're going to showcase performance in the upcoming event next week.

Can't wait to see it honestly!

I am really tired of companies hyping things up without any real meat to the info they're giving.

they all do it and it doesn't always turn out favorably for the company.

I am not being specifically critical on AMD either, Nvidia and Intel and countless game studios have done the same thing.

Hype is a good thing but we want information that's all

[TechFan@ic]

Can't wait to see it honestly!

I am really tired of companies hyping things up without any real meat to the info they're giving.

they all do it and it doesn't always turn out favorably for the company.

I am not being specifically critical on AMD either, Nvidia and Intel and countless game studios have done the same thing.

Hype is a good thing but we want information that's all

AMD did not hype anything up, they didn't even claim any specific performance numbers, today's announcement was just about more partnerships.

[beebskadoo]

AMD did not hype anything up, they didn't even claim any specific performance numbers, today's announcement was just about more partnerships.

Ok, fair enough.

The hype that's generated is mostly from the Developers and from people that are quick to give it a thumbs up without seeing the performance numbers.

AMD knows this and yet nothing has been released. IDK if that's a good thing or a bad thing, hopefully a good thing but it's annoying.

I just hate it when companies launch a product with the promise of future features that they don't give any info on. Nvidia did this with the Shield, and to an extent Kepler, and it makes me shake my head because it's like DLC for a game. Why not have these things ready at launch rather than months down the road. I understand implementing new tech down the road but promising something right before the launch of a card and expecting people to buy based on a promise is a bad strategy.

just my thoughts on the subject I suppose

[TheMissxu]

I agree marketing products with Mantle isn't the most postitively moral thing to do, but we should have no problem with extra features added to a product after you’ve bought it (That's a good thing even if the stuff is horrid).

Linus did buy a camera with the hopes for 4k support later down the line. (Just food for thought.)

[Whiskers]

Until I see real performance boosts, I'm not interested in their PR rubbish.

[Vitalius]

That doesn't change anything you still have to make the game for Mantle to get a peformance gain wich no dev will do unless AMD gives them money to do so.

(Like Nvidia does with PhysX)

*facepalm*

That changes everything. 

The reason people have to be paid to program for PhysX is because it only applies to some, but not all, of the market. That doesn't apply for Mantle. If a game is purely developed in Mantle, it should work with DirectX by default. 

Everyone uses DirectX. So that means they don't have to program in DirectX at all. Which means it takes the same amount of resources as choosing DirectX, except that choosing Mantle gives a performance boost with AMD GPUs, but works with Nvidia GPUs because it works with DirectX. The only exception to that is that it will cost more time and money to learn Mantle, but at the same time, that is somewhat worth it imo.

That's completely different from having to program a game in both DirectX and Mantle, as that literally doubles the cost, then adds more because they need to learn Mantle.

That's the implication. I could be wrong, but that is what AMD's press releases have implied. 

[JayD]

I really hope Mantle takes off and lives up to all the hype so that more developers implement it.