Oxide Games Says AMD Mantle is Disruptive Technology

[gunbust3r]

Designing a ham sandwich that costs $20 with PCIe and an Displayport connector that renders 120FPS at 4K would be disruptive. Making your $600 video card ~25% faster not so much...

 

[vampyiere6]

I dont get why ppl are hating on mantle its a very good thing even when its still in "beta" so i think for the comming years it will still be better and most gaming companies will use it. and ppl hating because nvidia wont benefit from it since its amd gpu's only its nvidia that dont want it because they want their own stuff that u have to pay premium price for that wont benefit you. mantle is free for them but nvidia dont wanna use it in their cards. sorry for my bad english

 

[Jaroslav Jandek]

Jaroslav : are you daydreaming?? If no one suppors mantle because it involves additional work, it won't caught up. No other new API will never unless some labor is put into it. Currently mantle is in its infancy, lets wait to see how it develops before trying to kill it please.

I am not daydreaming, I am simply stating facts. There is no reason for me (an independent dev) to develop a module that will require man-hours, product support and will marginally benefit only a small fraction of customers. There are big companies already supporting it and maybe over time it will gain traction. I am NOT trying to kill it, I am just providing a different perspective.
I actually think Mantle has potential but I very much dislike the PR stunts surrounding it.

 

[bloodroses75]

Microsoft will try to find a way to kill Mantle, just like they did to SDL. They don't like competition of any form.

 

[TheAshigaru]

For all of you who are upset about the word disruptive, from wikipedia:A disruptive innovation is an innovation that helps create a new market and value network, and eventually disrupts an existing market and value network (over a few years or decades), displacing an earlier technology. The term is used in business and technology literature to describe innovations that improve a product or service in ways that the market does not expect.

Technically you are correct, but it still was a bit misleading. 'Disruptive' has a negative connotation associated with it. The main problem is the location of the information in the article. To keep the title as is, the first paragraph should have explained how Mantle is a 'disruptive technology.' Instead we have to read through to the last paragraph. Even then, it doesn't really give a great explanation of how Mantle is disruptive--just a link. It doesn't really follow typical journalistic style.On the other hand, it still was pretty informative. It's nice to get a better explanation of what AMD meant by 'close to the metal' without all the marketing jargon.

 

[tomc100]

The problem with mantle is that it only works with a GCN architecture. On a console where the gpu doesn't change it makes sense for close to the metal API but on a pc where architecture can and will change then newer technology or older technology can't use it. The possible good thing about mantle is that it will provide competition with directx so Microsoft needs to improve that API or not restrict future updates to windows 8 in order to sell their os. Also, it can lower hardware specs for steam machines if a game uses it to reduce cost. Right now, steam machines don't make any sense unless you can reduce the price by half and still get a machine that can play games at high fidelity.

 

[CaedenV]

What is with the stupidity of people today? When has disruptive technology ever had negative connotations in the tech industry (or most industries for that matter)? Why would anyone read this title and think it is a bad thing?Disruptive technology is technology that fundamentally changes the way an industry works because it is fundamentally easier or better to use. It is disruptive because it disrupts the normal pattern of business to a point where it is impractical (if not impossible) to go back. No negative connotations here, just an adjustment period as people learn to use it better to their advantage.I still dobut that Mantle is going to be any more disruptive than CUDA or PhysX was for nVidia. CUDA was/is very useful for some things, and PhysX is very neat in some situations, but neither are transformational or disruptive. They are nice added bonuses when they happen to be supported, but hardly 'must have' features for any piece of software to be functional or useful (except in rare specialized cases that only a handful of people care about).

 

[calken]

I think this is more of a clever strategic marketing plan by AMD.a) Introduce a new API which allows your title to have better framerates, better gfxb) Software houses push the limit to make their title look more impressivec) Consumers buying new cards choose AMD to get higher framerates and better gfx on AAA titled) NVidia and Intel join Mantle to to gain salese) Smaller houses move exclusively to Mantlef) Junior devs gain opportunity of learning Mantle and becoming attractive to software housesg) Consumers benefit from better quality gfx on cheaper gfx cardsh) SteamOS becomes awesomei) Linux gets even betterj) SteamOS gets ported to mobile OS, phone becomes your SteamBoxh) x86 dies and gets replaced by ARMi) ARM buys AMDj) Apple implodes because Johnny Ive leaves(I'm sure I started off with just 3 points.....)

 

[sugetsu]

misleading title much?he says its the (disruptive) technology that would get everyone to rethink thingsthe way the title puts it is more to the effect that he said its a bad technology - the complete opposite as to what he actually said

Indeed, I just don't get why hasn't this website even bother running some mantle benchmarks and analysis in the first place. While every other tech website is raving about mantle and its potential, Tom's hardware has been conspicuously muted on the matter.

 

[internetlad]

I think the issue here is not with the actual definition of the word, it's with the journalistic definition of the world. Technically you see headlines like "Small town decimated by bomb". . . but you never actually get into the article and see that it reduced the town by 1/10th it's actual size. The word "Disruptive" has negative journalistic connotations, and the title is very misleading to people who are glancing over the daily headlines. Whether this was the author's intention or not, it's very poor journalism.

 

[Haravikk]

The problem with Mantle IMO is that it's AMD only; Direct 3D may be Microsoft only, but at least it's supported by all the graphics cards your games are likely to run on, but what's going to happen with Mantle when it comes to NVidia cards? I can see many developers just sticking to Direct 3D only because it supports all of their target market.The other problem is that AMD seems to be neglecting Direct 3D performance, in particular multithreaded workloads, which means that Mantle potentially only has the advantage because these features aren't implemented for the competition.I dunno, I like the idea of Mantle in theory, but I'd be much happier if it were actually some all new version of OpenGL that all cards would support, and support well.

 

[Djentleman]

@Haravikk Get ready to get down voted. People don't like hearing reality.

 

[vampyiere6]

The problem with Mantle IMO is that it's AMD only; Direct 3D may be Microsoft only, but at least it's supported by all the graphics cards your games are likely to run on, but what's going to happen with Mantle when it comes to NVidia cards? I can see many developers just sticking to Direct 3D only because it supports all of their target market.The other problem is that AMD seems to be neglecting Direct 3D performance, in particular multithreaded workloads, which means that Mantle potentially only has the advantage because these features aren't implemented for the competition.I dunno, I like the idea of Mantle in theory, but I'd be much happier if it were actually some all new version of OpenGL that all cards would support, and support well.

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

 

[renz496]

The problem with Mantle IMO is that it's AMD only; Direct 3D may be Microsoft only, but at least it's supported by all the graphics cards your games are likely to run on, but what's going to happen with Mantle when it comes to NVidia cards? I can see many developers just sticking to Direct 3D only because it supports all of their target market.The other problem is that AMD seems to be neglecting Direct 3D performance, in particular multithreaded workloads, which means that Mantle potentially only has the advantage because these features aren't implemented for the competition.I dunno, I like the idea of Mantle in theory, but I'd be much happier if it were actually some all new version of OpenGL that all cards would support, and support well.

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

AMD might say it is open for other to use it but i think even AMD is well aware than nvidia and intel might have no interest to pick up Mantle if they keep the API development to themselves. the very same reason why AMD refuse the offer from nvidia to license PhysX so they can implement GPU PhysX on their GPU. on developer side game dev have no good reason to code for mantle if Mantle are not adopted by other gpu vendor. i'm sure AMD know about this that's why they say others can also use Mantle if they want to but so far they leave it at that and never give more details about it.

as for nvidia making their own version of mantle that's remain to be seen. if mantle really poses the threat then they might. but so far they push DirectX/OpenGL more than AMD. some people complaining that nvidia hardware like Kepler is not compliant with DX11.1 and later specification but in reality it is nvidia that really use the feature for DX11 while AMD decides not to implement it in their drivers

 

[Haravikk]

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

As others have said, while it's nice to see it open, it's still AMD's. Don't get me wrong, I like AMD, but this is something that should be really be run by its own, vendor neutral open source foundation, so that different vendors are encouraged to contribute.

That said, for something entirely new it's going to be an uphill struggle, meanwhile OpenGL is in that position now; granted it has had numerous problems of its own, but I think that more realistically they're the ones that could actually push a Mantle-like API into being adopted. The problem is whether they can, given the issues with changing OpenGL in the past, but I'm hoping it's something that could happen as there's now a lot of incentive for Valve to push OpenGL forwards since it's the only real API of choice on Linux. Besides, what's good for OpenGL is good for everyone IMO, since it's supported by all major platforms, and all major hardware, already.

 

[Mousemonkey]

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

If Mantle is as "Open" as you seem to think it is then why is that AMD have not seen fit to demonstrate to the world that it can run on a non GCN architecture? This they could easily do and still keep it proprietary (if they wish to) by enabling it to run on the 6xxx and 5xxx series cards.

 

[childofthekorn]

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

If Mantle is as "Open" as you seem to think it is then why is that AMD have not seen fit to demonstrate to the world that it can run on a non GCN architecture? This they could easily do and still keep it proprietary (if they wish to) by enabling it to run on the 6xxx and 5xxx series cards.

Vampyiere6 is mistaken. It will be open in the future but currently, as it is still in beta, they declare it proprietary. Once it is released with a larger amount of features, AMD has stated they will be opening it up.

You mention that its not Open because they tailored it to use the GCN architecture...ok, well...Nvidia can actually make it use their architecture once it is opened up, just as AMD has done for theirs. You can't expect one company to work on making the product work on their competitors products, thats not AMD's responsibility. Nvidia didn't do so with GSync, Physx and their other proprietary offerings, why should AMD? The fact that AMD has given so much to the Open Source community should entitle them to at least finish production on a piece of software before releasing the source code to their competitors. Its still in BETA and is projected to be in this state until ~2015.

 

[childofthekorn]

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

As others have said, while it's nice to see it open, it's still AMD's. Don't get me wrong, I like AMD, but this is something that should be really be run by its own, vendor neutral open source foundation, so that different vendors are encouraged to contribute.

That said, for something entirely new it's going to be an uphill struggle, meanwhile OpenGL is in that position now; granted it has had numerous problems of its own, but I think that more realistically they're the ones that could actually push a Mantle-like API into being adopted. The problem is whether they can, given the issues with changing OpenGL in the past, but I'm hoping it's something that could happen as there's now a lot of incentive for Valve to push OpenGL forwards since it's the only real API of choice on Linux. Besides, what's good for OpenGL is good for everyone IMO, since it's supported by all major platforms, and all major hardware, already.

Vamp was wrong, its proprietary. AMD has stated they will be releasing the code so others can contribute once it exits its BETA stage. That is projected to be sometime in ~2015.

 

[Mousemonkey]

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

If Mantle is as "Open" as you seem to think it is then why is that AMD have not seen fit to demonstrate to the world that it can run on a non GCN architecture? This they could easily do and still keep it proprietary (if they wish to) by enabling it to run on the 6xxx and 5xxx series cards.

Vampyiere6 is mistaken. It will be open in the future but currently, as it is still in beta, they declare it proprietary. Once it is released with a larger amount of features, AMD has stated they will be opening it up.

You mention that its not Open because they tailored it to use the GCN architecture...ok, well...Nvidia can actually make it use their architecture once it is opened up, just as AMD has done for theirs. You can't expect one company to work on making the product work on their competitors products, thats not AMD's responsibility. Nvidia didn't do so with GSync, Physx and their other proprietary offerings, why should AMD? The fact that AMD has given so much to the Open Source community should entitle them to at least finish production on a piece of software before releasing the source code to their competitors. Its still in BETA and is projected to be in this state until ~2015.

That remains to be seen and as such is still speculative.

 

[Haravikk]

Vamp was wrong, its proprietary. AMD has stated they will be releasing the code so others can contribute once it exits its BETA stage. That is projected to be sometime in ~2015.

Thanks for clarifying, still though there are too many issues surrounding it; thus far it seems AMD is intentionally ignoring Direct 3D improvements in order to give Mantle such a large competitive edge, we have no idea how big an improvement we would see on hardware that has had just as much Direct 3D Optimisation. This means that they're almost in a position where they're trying to force developers to switch from Direct 3D, but if continuing in Direct 3D is how to ensure support for NVidia too, then I don't see developers shifting.

It's admirable that they want to make it open source, but it's still very much AMD branded, which is why I was saying it would make a lot more sense for this to be an independent open source effort, even if it was still backed by AMD as the primary contributor. That's what OpenGL should be, but given the difficulties in revolutionising their API in the past it's hard to say that Mantle should be an OpenGL release, but at least if it were an OpenGL technology then NVidia would almost certainly support it. The best example I can think of is CUDA and OpenCL; CUDA is a good technology, but AMD were never going to support it and Intel doesn't either, but the hardware from all three companies supports OpenCL, which is likely to become the only real compute technology going forward once current CUDA software makes the transition. Anything being too proprietary introduces big potential barriers here, even if the only proprietary part is branding, but for now we only have AMD's word for it.

I still think Valve may have a big role to play here; if they're happy with what Mantle's open-source credentials are, or will be, then they might back it, and find a way to encourage NVidia to take it on board as well. Otherwise Valve is going to have to back OpenGL, and try to push that to make similar improvements, or introduce new alternate APIs that can compete with Mantle, otherwise Steam OS is potentially going to be at a big performance advantage for some time.

 

[childofthekorn]

The problem is that its all open but nvida dont want it cost they wanna use their own stuff that they can put a premium price on, nvida only make stuffs that only works for their own cards. i dont say amd dosent do it but they are much more open and use stuff everyone can use if they want.

If Mantle is as "Open" as you seem to think it is then why is that AMD have not seen fit to demonstrate to the world that it can run on a non GCN architecture? This they could easily do and still keep it proprietary (if they wish to) by enabling it to run on the 6xxx and 5xxx series cards.

Vampyiere6 is mistaken. It will be open in the future but currently, as it is still in beta, they declare it proprietary. Once it is released with a larger amount of features, AMD has stated they will be opening it up.

You mention that its not Open because they tailored it to use the GCN architecture...ok, well...Nvidia can actually make it use their architecture once it is opened up, just as AMD has done for theirs. You can't expect one company to work on making the product work on their competitors products, thats not AMD's responsibility. Nvidia didn't do so with GSync, Physx and their other proprietary offerings, why should AMD? The fact that AMD has given so much to the Open Source community should entitle them to at least finish production on a piece of software before releasing the source code to their competitors. Its still in BETA and is projected to be in this state until ~2015.

That remains to be seen and as such is still speculative.

“The plan is, long term, once we have developed Mantle into a state where it’s stable and in a state where it can be shared openly [we will make it available]. The long term plan is to share and create the spec and SDK and make it widely available. Our thinking is: there’s nothing that says that someone else could develop their own version of Mantle and mirror what we’ve done in how to access the lower levels of their own silicon. I think what it does is it forges the way, the easiest way,” explained Mr. Corpus.

http://www.xbitlabs.com/news/graphics/display/20131003232053_AMD_Mantle_Is_Proprietary_Now_But_It_Will_Become_Widely_Available_t😵thers.html

Although yes, things may change in the year leading up to the release, they've gone on record to show that they intend to release it once they work out the kinks. Hopefully sooner than later. I'd love to see the same API embraced by every hardware manufacturer as a benefit to the consumer, in an open source setting of course, to ensure great performance, or even far more innovation in games with the allowance of more code to be added to a developers game.

 

[Mousemonkey]

It's interesting that you post a link that I also posted way back in the early Mantle threads! :lol:

 

[childofthekorn]

It's interesting that you post a link that I also posted way back in the early Mantle threads! :lol:

The joys of instant information 😛

 

[somebodyspecial]

So Oxide, who was paid to promote Mantle on a tech demo says it didn't put much effort into it. I'll take that with a grain of salt...

My thoughts exactly. I want to hear a guy who HAS NOT received a single penny from AMD and has used mantle. I want to see THAT guy claiming it's great and for the love of god I want to see someone stating we got X percent gain over directx. I do not want magical quotes with no REAL numbers, and even then saying we have no ability to test actual speeds, someone bigger can do that. Translation, we're too afraid to tell actual numbers we are getting or we'd be bragging right and left with percentages and REAL NUMBERS. All I've seen so far is comments without data to support it. Much like the claims saying AMD is selling out cards like crazy to miners, when every card is in stock and AMD made ZERO gpu money on this last quarter when supposedly, all this mining was going on.

The only data we do have from BF4 shows you get less than 10% on high end hardware (even from AMD loving anandtech site it was 7-10%) which for most gamers would mean it's almost pointless. A driver update can get that and affect tons of games. You have to wait for games to USE mantle and they may never use it. Tomshardware decided to leave MANTLE out of the results with AMD's 265 due to the poor showing of the BETA API. So pretty clear my statements are correct at least for now or toms would publish the data and everyone else would be giving actual DATA percentages not "it's really good"...LOL. It MIGHT be one day, but it certainly isn't now.

 

[childofthekorn]

So Oxide, who was paid to promote Mantle on a tech demo says it didn't put much effort into it. I'll take that with a grain of salt...

My thoughts exactly. I want to hear a guy who HAS NOT received a single penny from AMD and has used mantle. I want to see THAT guy claiming it's great and for the love of god I want to see someone stating we got X percent gain over directx. I do not want magical quotes with no REAL numbers, and even then saying we have no ability to test actual speeds, someone bigger can do that. Translation, we're too afraid to tell actual numbers we are getting or we'd be bragging right and left with percentages and REAL NUMBERS. All I've seen so far is comments without data to support it. Much like the claims saying AMD is selling out cards like crazy to miners, when every card is in stock and AMD made ZERO gpu money on this last quarter when supposedly, all this mining was going on.

The only data we do have from BF4 shows you get less than 10% on high end hardware (even from AMD loving anandtech site it was 7-10%) which for most gamers would mean it's almost pointless. A driver update can get that and affect tons of games. You have to wait for games to USE mantle and they may never use it. Tomshardware decided to leave MANTLE out of the results with AMD's 265 due to the poor showing of the BETA API. So pretty clear my statements are correct at least for now or toms would publish the data and everyone else would be giving actual DATA percentages not "it's really good"...LOL. It MIGHT be one day, but it certainly isn't now.

Although you make good points, considering AMD's financial situation the last 10 years (or so) they could produce only 15 cards and sell them and that statement would be true, including the financial toll they took from GPU's that quarter. I'm personally very hopeful for mantle, from a consumer standpoint. However, I agree with you that I would like to see many more scenarios from a 3rd party to narrow down which areas it'll excel in. Especially once it matures.