Firaxis discusses why Mantle is revolutionary for their work with Civilization: Beyond Earth

[opethdisciple]

They are being paid to say all of this by AMD. They have even been paid to use it.

DX12 will make Mantle redundant. At the moment AMD are pushing it as a technology advantage in order to generate sales.

That’s not to say it isn’t useful presently.

/thread

 

[Sugarhell]

They are being paid to say all of this by AMD. They have even been paid to use it.

DX12 will make Mantle redundant. At the moment AMD are pushing it as a technology advantage in order to generate sales.

That’s not to say it isn’t useful presently.

/thread

Did you even read the thread before you call it /thread? They said that they spent money to support it but its worth it for the future.Mantle is perfect for RTS.

 

[opethdisciple]

Did you even read the thread before you call it /thread? They said that they spent money to support it but its worth it for the future.Mantle is perfect for RTS.

In fairness I did not. And then when I reread my post, I admit it was a little bit unneccesary.

 

[drunkenmaster]

AMD paid DICE $10m to help develop Mantle, But they can't afford to keep paying for people to use it.

No, they didn't pay Dice 10mil for that, AMD/Nvidia have paid game dev's for exclusives(in terms of advertising/support gaming evolved/twimtbp) for pretty much a decade or more now. The $10mil is for the exclusive and all the copies of the game AMD bundled.

Bundled games AMD/Nvidia include with graphics are shockingly...... not free. AMD pays for them, Nvidia paid $6mil for a couple MUCH smaller ubisoft titles.

Mantle had nothing to do with the cost and Dice did not help develop it in a professional sense.

AMD made Mantle, it's that simple, as with DX, opengl and every other API those using it were included in discussions. Johan was a big supporter of a new API and jumped in helping on Mantle because he wanted to.

Pretend Mantle doesn't exist, AMD or Nvidia would still have paid a similar amount to win the advertising and game bundles exclusive.

 

[PhysicsMan94]

Whilst I agree with the programming principles behind Mantle I absolutely do not support Mantle itself due to the closed nature of the API. It is exclusive to AMD customers simply because Mantle has been coded specifically for the GCN architecture, that is the whole point - DirectX is not as efficient because it is not a graphics API built for a specific hardware architecture, given that DX developers must program for a whole wide range of hundreds of different hardware configurations it is actually impressive the performance one can squeeze out of this API.

Mantle is a one-trick pony and it just acts to regress the graphics and gaming industry down a path of exclusivity and all the BS that comes with that. The technology itself is progress, but the very nature of the API means that it is an exclusive feature for AMD users - this is not good and it's the very thing that I thought lots of people on this forum stood against.

In my opinion, it would've been better if AMD worked more closely with Microsoft and Nvidia in developing DirectX to solve the problems of generalised hardware optimisations - all 3 companies are very wealthy and have lots of incredibly talented people working for them, together they would've come up with a DirectX based solution which would likely give performance near to Mantle (but available for all configurations). Instead, AMD walked the selfish path and decided to take the dangerous, political road to a closed API which will just act to fragment the gaming and graphics industries even further - absolutely not what is needed.

AMD need to be careful, if they continue down this path and try to abandon DirectX whilst pushing their closed API "Mantle" then Microsoft will just make sure that they optimise DX more for Nvidia cards so that DirectX based games and other applications will perform significantly better than AMD cards on DirectX.

 

[Final8y]

AMD need to be careful, if they continue down this path and try to abandon DirectX whilst pushing their closed API "Mantle" then Microsoft will just make sure that they optimise DX more for Nvidia cards so that DirectX based games and other applications will perform significantly better than AMD cards on DirectX.

Where has AMD or anyone involved suggested that are trying to abandon DX, got a link to that ?

 

[PhysicsMan94]

Where has AMD or anyone involved suggested are trying to abandon DX.

I'm not saying they have said that, but it is an option which they could conceivably take if Mantle takes off (which I am doubtful of fortunately).

 

[jh30uk]

I am pretty sure from reading the past 6+ months of mantle propaganda that they though it was a DX killer.

 

[Orangey]

I doubt Mantle will replace DX in most games, most developers need the hand-holding or simply don't need Mantle.

Most likely DX12 will be 10% slower than Mantle or somesuch, then Crytek et al who need to grub every last CPU cycle will be the ones using Mantle.

 

[Final8y]

I'm not saying they have said that, but it is an option which they could conceivably take if Mantle takes off (which I am doubtful of fortunately).

I am pretty sure from reading the past 6+ months of mantle propaganda that they though it was a DX killer.

As the context as in better which does not mean wholeheartedly replace when the conditions are nowhere near or even likely ever will be, even if the Mantle path was available in every title the conditions would not be there to replace DX.

Mantle would have to be running on NV and Intel GPUs for that to be even a twilight thought.

 

[tommybhoy]

Good info matt.

That line is a bit worrying, are they saying that they wont be doing the full rendering pipeline that DirectX will need?
Or as I hope that they are saying, 'this extra rendering pipeline that we are doing is less complex than the one for DirectX that we already have.'

At present, the benefits of Mantle extend only to those customers which can run it. We recognize that a large fraction of our customers will not have access to Mantle, and we do not intend to discriminate.
Our philosophy is to strive to use our customers’ machines to their fullest potential. To the extent possible, DirectX customers will see the same images as Mantle customers, and we will provide DirectX customers with the highest performance that their systems are capable of. It is precisely this motivation which impels us to offer Mantle to those customers who can use it, because their machines possess great untapped potential. By tapping that potential, we hope to drive positive changes which will eventually spread to all of our other customers.

 

[Final8y]

I doubt Mantle will replace DX in most games, most developers need the hand-holding or simply don't need Mantle.

Most likely DX12 will be 10% slower than Mantle or somesuch, then Crytek et al who need to grub every last CPU cycle will be the ones using Mantle.

It should not replace DX in any games as in the only option while the competition cant run it, that would be silly and i for one would not be cheering about it If a Mantle only game came out under the current conditions.

 

[drunkenmaster]

The problem with AMD asking MS to work on a new version of DX is, when adding more people you get more opinions and instantly a longer time to market. It wasn't in any way in MS's interest to enabled PC to be even faster than consoles, remember consoles have had this low API overhead for donkeys years and MS knew categorically how much it improved the situation yet did nothing.

AMD could get a low level API to market massively quicker effectively on it's own without involving anyone else and FORCE the rest of the industry to either jump on board or make a competitive API. It succeeded and very very quickly to attain the result that was intended.

Now imagine MS was on board, and they wanted everyone else on board, put few people on board, kept wanting to meet up to discuss things, votes on who to involve, votes and discussions on every decision big and small. It could have taken much much longer. Now that Mantle is out there, proving it's advantages to the degree that MS on knowing it was coming kick started DX12 into full gear. Could be 20 times as many people involved now than if MS was involved from the start and trying their best to delay it.

When it comes to new standards and new idea's, they almost universally come around faster when someone just does it and everyone else plays catch up.

For me AMD did the absolutely right thing with Mantle, if it hadn't happened this way I think early 2016 for DX12 games would be a pipedream right now.

 

[Orangey]

The problem with AMD asking MS to work on a new version of DX is, when adding more people you get more opinions and instantly a longer time to market. It wasn't in any way in MS's interest to enabled PC to be even faster than consoles, remember consoles have had this low API overhead for donkeys years and MS knew categorically how much it improved the situation yet did nothing.

AMD could get a low level API to market massively quicker effectively on it's own without involving anyone else and FORCE the rest of the industry to either jump on board or make a competitive API. It succeeded and very very quickly to attain the result that was intended.

Now imagine MS was on board, and they wanted everyone else on board, put few people on board, kept wanting to meet up to discuss things, votes on who to involve, votes and discussions on every decision big and small. It could have taken much much longer. Now that Mantle is out there, proving it's advantages to the degree that MS on knowing it was coming kick started DX12 into full gear. Could be 20 times as many people involved now than if MS was involved from the start and trying their best to delay it.

When it comes to new standards and new idea's, they almost universally come around faster when someone just does it and everyone else plays catch up.

For me AMD did the absolutely right thing with Mantle, if it hadn't happened this way I think early 2016 for DX12 games would be a pipedream right now.

Yup, funny how MS started shouting about DX12 after half a decade literally AS SOON AS Mantle came out. The joint propaganda barrage with NV is pretty funny too, with NV claiming they "worked on it with MS" for like 5 years as both companies need to look like they weren't just twiddling their thumbs the whole time. In reality I think the project probably started the day of AMD's Hawaii event and was cobbled together from ideas researchers had been playing with in their spare time.

 

[GoogalyMoogaly]

DirectX, or at least DX11 is very difficult, time consuming and clumsy to use, its expensive Developing games for DirectX

Mantle is nice, simple and cheap, what i would like to see is for Nvidia to make their own Mantle.

Tell that to the people that made Thief complete with no crossfire Mantle support! And yet DirectX does have multi-card support?

 

[TheRealDeal]

Tell that to the people that made Thief complete with no crossfire Mantle support! And yet DirectX does have multi-card support?

You then have to factor in how long was spent on dx and how many people were focused on dx v the same for Mantle. Those with single gpu's run in Mantle because the experience is better than DX. Factor all that in and you have your answer. DX is still the focus and these developers are saying it's only 1 guy or a few guys integrating mantle into the game over 3 month's. Imagine how good games would be if Mantle was the main Focus and how bad DX would be if this were the other way round. It would really make DX look stupid and out dated is my bet.

 

[GoogalyMoogaly]

You then have to factor in how long was spent on dx and how many people were focused on dx v the same for Mantle. Those with single gpu's run in Mantle because the experience is better than DX. Factor all that in and you have your answer.

But Mantle is so quick and easy to do. BF4 took like 4 hours with 1 guy doing it while walking his dog (or something).

And of course there is another point there. Why spend so much time and effort on DirectX if it's so bad? Maybe because Mantle has such a limited audience?
If AMD are so good, why didn't they produce an API that would run on all cards? GCN, non-GCN, Nvidia, Intel, like Microsoft is?
Can the API really catch on while supporting such a small number of cards?

Don't they realise that games can't really change while they all have to be playable on DirectX?
How different will "Civ6" be when run on Mantle considering the game will need to run on DirectX11 machines too and there are a ****-ton more DirectX 11 system than Mantle systems/customers?
I think, in time that's where DX12 could change things while Mantle never will (unless Nvidia and maybe Intel get onboard, but DX12 make that seem less likely). Once DX12 really takes hold games can (hopefully) be designed differently, with the knowledge that they have more CPU resource to use. Maybe then we'll see something more than slightly nicer graphics effects and higher framerate (as nice as they are) as was demo'd with StarSwarm.

I'm not really fussed either way with Mantle, if it continues for years and years (until we're complaining about that too) or if it dies off once DX12 takes hold.
I do hope DX12 does at least basically for everyone what Mantle has done for a small minority.
I'm also looking forward to seeing a new thread for every DirectX game that is announced (and probably another when it's released) until they replace it with something else...

 

[rolypolyman]

Yet to taste the benefits of 'Mantle' despite the fact I have an R290, don't play BF4 and Thief was a huge disappointment AND most importantly have yet to get WHQL drivers which do the job. Starting to think that by the time Mantle is 'ready' I wont care, oh and I have no more faith in Microsoft getting it right with DX12.

 

[TheRealDeal]

But Mantle is so quick and easy to do. BF4 took like 4 hours with 1 guy doing it while walking his dog (or something).

And of course there is another point there. Why spend so much time and effort on DirectX if it's so bad? Maybe because Mantle has such a limited audience?
If AMD are so good, why didn't they produce an API that would run on all cards? GCN, non-GCN, Nvidia, Intel, like Microsoft is?
Can the API really catch on while supporting such a small number of cards?

Don't they realise that games can't really change while they all have to be playable on DirectX?
How different will "Civ6" be when run on Mantle considering the game will need to run on DirectX11 machines too and there are a ****-ton more DirectX 11 system than Mantle systems/customers?
I think, in time that's where DX12 could change things while Mantle never will (unless Nvidia and maybe Intel get onboard, but DX12 make that seem less likely). Once DX12 really takes hold games can (hopefully) be designed differently, with the knowledge that they have more CPU resource to use. Maybe then we'll see something more than slightly nicer graphics effects and higher framerate (as nice as they are) as was demo'd with StarSwarm.

I'm not really fussed either way with Mantle, if it continues for years and years (until we're complaining about that too) or if it dies off once DX12 takes hold.
I do hope DX12 does at least basically for everyone what Mantle has done for a small minority.
I'm also looking forward to seeing a new thread for every DirectX game that is announced (and probably another when it's released) until they replace it with something else...

Was Rome built in one day? No you say. Everything has to start somewhere. I think you are expecting far to much to fast.

 

[GoogalyMoogaly]

Was Rome built in one day? No you say. Everything has to start somewhere. I think you are expecting far to much to fast.

From what I hear I think Rome 2: Total War might have been knocked up in a day, does that count?

I'm still not convinced that it'll gain Nvidia and Intel support by the time DirectX 12 is out.