Hermits, did you guys know this all along? (Nvidia J'accuse!)

It failed hard before February 19th 2015, it failed before before July 2014 when Khronos started their project for a next gen API.

http://www.fool.com/investing/general/2014/03/25/its-official-amds-mantle-is-doomed.aspx

http://www.dsogaming.com/news/nvidia-finally-officially-speaks-about-amds-mantle-will-not-support-it-no-real-benefit-using-it/

A year before Vulcan announcment, AMD struggling to get developer support, DX12 announcement, OpenGL getting similar improvements. It may not have been "dead" a year before the announcement, but it was definitely circling the drain.

One reason I have never bought Nvidia hardware for PC, they have always been a bunch of dicks just like Intel.

Anti consumer tactics don't work on me.

LOL, Guess who has the most mature DirectX12 driver? It's not NVIDIA.

The link http://www.fool.com/investing/general/2014/03/25/its-official-amds-mantle-is-doomed.aspx refers to DirectX12. it's programming model similar to Mantle LOL.

www.fool.com's OpenGL remark was referring to NVIDIA's vendor specific OpenGL kit-bash which has been thrown out.

Mantle has serves it's purpose to change the industry towards AMD's programming model which is mostly for AMD's CPU interest.

Btw, NVIDIA's vendor specific OpenGL kit-bash deals with draw calls overheads.

Ok. Do you remember what this argument is about? Clearly you do not

@04dcarraher: lol making excuses..... AMD's Mantle was proprietary.... Them using Mantle as a base for Vulkan has no bearing on what AMD did before

AMD had a proprietary API they planned make public. It was not open, it was theirs. They were going to freely allow people to use the technology like their other technologies but but at the end of the day, Intel/NVidia have absolutely no say in the matter. It failed miserably. Then they transferred it to Khronos as a new API for OpenGL.

You are really bad at arguing, you very rarely keep things in the context of the argument, and more so than not you just post a blurb of facts that without putting them into context. The fact is 04dcarraher is correct. Your not even refuting the point any more. Do you even understand why DX12 is important here? It's announcement lost AMD some leverage with it's technology gaining traction. Do even understand why only 12 games matter.. it shows support for the technology. It wasn't much. But you point out I made an error with the companies involved EA, RSI, then I missed Square and who ever made sniper elite. The point is actually worse now because out of the publishes that picked up the technology, they really didn't take the tech very far if at all. The two I missed essentially dropped support for it immediately. NVIDIA's OpenGL kit-bash was important because it showed a possible alternative with similar results. Again losing some of mantle's value for developer adoption. Do you understand why it had to go to Khronos, because no one was going to support it. Months before Khronos announced it's plans for a new API, mantle was struggling to gain support, AMD's plan for a public API (which wasn't public as of yet) was looking worse with more appropriate technologies that were more impartial to hardware vendors.

If you go back far enough in this thread, You'll see I actually argued against 04dcarraher originally. I thought it was an open api well before any dates involving the khronos group. It was not. AMD clearly stated open when it was a proprietary API they owned but were going to allow people to use freely without restriction at some point. Is that a nice thing for them to do. Yes. But is it also misleading when they call it open? Yes, they owned it, they had the final say in how it developed, ad the openness is completely subject to change in later iterations if they want.

I understand your arguments, the technoligy basically still exists, DX12 is similar, vulkan is a continuation of mantle. Ok. That does not refute or prove the core issue we have been discussing. Very little of what you said actually does. DX12 and Vulkan are still, not mantle, mantle ended at 1.0. They may be similar or share a lot in common with. But mantle is dead.

@waahahah:

Again,

@04dcarraher: lol making excuses..... AMD's Mantle was proprietary.... Them using Mantle as a base for Vulkan has no bearing on what AMD did before

Wrong. A view from Unity3D's Aras Pranckevičius http://aras-p.info/blog/2015/03/13/thoughts-on-explicit-graphics-apis/

Khronos Vulkan was announced (which is very much AMD Mantle, improved to make it cross-vendor).

From Unity's programmers POV, "Vulkan has no bearing on what AMD did before" assertion is false.

What ever knowledge learnt from Mantle, they are being transferred to Vulkan.

@waahahah:

AMD had a proprietary API they planned make public. It was not open, it was theirs. They were going to freely allow people to use the technology like their other technologies but but at the end of the day, Intel/NVidia have absolutely no say in the matter. It failed miserably. Then they transferred it to Khronos as a new API for OpenGL.

The problem, AMD's other technologies was given to open standards bodies e.g. HBM (JEDEC), GDDRx (JEDEC), FreeSync (VESA), HSA.

Against NVIDIA's Gameworks, AMD's GPUOpen has open source MIT license.

"Failing miserably" has nothing to do with keeping Mantle API spec to themselves. What's important is giving the proprietary API to open standards bodies.

AMD could have elected to not give Mantle API spec to Khronos and remain with legacy OpenGL. AMD's alternative plan for legacy OpenGL was to add Mantle like extensions (October 3, 2013).http://www.dsogaming.com/news/amd-aims-to-give-opengl-a-big-boost-api-wont-be-the-bottleneck/

DirectX12 has largely fulfilled Mantle's original goals to move the industry towards AMD's interest.

@waahahah:

NVIDIA's OpenGL kit-bash was important because it showed a possible alternative with similar results. Again losing some of mantle's value for developer adoption. Do you understand why it had to go to Khronos, because no one was going to support it. Months before Khronos announced it's plans for a new API, mantle was struggling to gain support, AMD's plan for a public API (which wasn't public as of yet) was looking worse with more appropriate technologies that were more impartial to hardware vendors.

NVIDIA's OpenGL kit-bash extensions developer's support was worst than Mantle. The dominant OpenGL GPU vendors are non-NVIDIA and that's on Android OS e.g. ARM, Qualcomm and Imagination Technologies. All 3 OpenGL GPU vendors dislikes NVIDIA's attempted fragmentation in the Android OS market.

You understand NVIDIA's OpenGL kit-bash extensions wasn't enough to negate December 3, 2014's Khronos Group's publically called for participants.

The majority of Khronos Group voted to dump the legacy OpenGL and select an API that was built for modern GPUs.

Mantle's value is to move the industry towards AMD's interest which is mostly the CPU side. If AMD has powerful CPU, these API changes would not be necessary.

Umm. Yeah. I don't care that much. If I did then I'd still be crying over 3dfx.

@ronvalencia:

why dont you stop it your wrong, fact is that Mantle wasn't used by any other gpu company, and to the fact only one architecture ie GCN was ever allowed to use it by the company that made the API....... it means that it was proprietary. the "what if intent" and later reincarnation of it means squat when the fact is that AMD's Mantle was only used by AMD products.....

One word 'overclocking'. In my book this is where AMD falls over, more than anything else.

My 980ti has a near 25% overclock when compared to the reference card. You're not getting that kind of OC on water, let alone on air, with AMD cards

Again this doesn't say anything about ownship, irrelavent.

From Unity's programmers POV, "Vulkan has no bearing on what AMD did before" assertion is false.

What ever knowledge learnt from Mantle, they are being transferred to Vulkan.

You're missing the point, of the statement in context. Of course what did did before matters, but we are strictly talking about the licensing not the technology.

The problem, AMD's other technologies was given to open standards bodies e.g. HBM (JEDEC), GDDRx (JEDEC), FreeSync (VESA), HSA.

Against NVIDIA's Gameworks, AMD's GPUOpen has open source MIT license.

Has no baring on the decisions made throughout mantle development. They explicitly stated time and time again to make it a public API. GPUOpen being MIT doesn't matter since its still trademarked through them and that license is subject to change. Which is why intel/nvidia probably won't adopt it. It's the same with any real open source project, the owner of the source code can change the licensing on future versions of the software.

"Failing miserably" has nothing to do with keeping Mantle API spec to themselves. What's important is giving the proprietary API to open standards bodies.

AMD could have elected to not give Mantle API spec to Khronos and remain with legacy OpenGL. AMD's alternative plan for legacy OpenGL was to add Mantle like extensions (October 3, 2013).http://www.dsogaming.com/news/amd-aims-to-give-opengl-a-big-boost-api-wont-be-the-bottleneck/

DirectX12 has largely fulfilled Mantle's original goals to move the industry towards AMD's interest.

You're link does not state AMD will give the API to khronos, but they will give similar extensions... do even read the things you post?

"AMD, the red team will be supporting this open API with some high performance extensions that will offer almost similar performance to AMD’s upcoming API, Mantle."

Again this is confirming AMD's plan NOT to give it away. It's not really helping your case.

NVIDIA's OpenGL kit-bash extensions developer's support was worst than Mantle. The dominant OpenGL GPU vendors are non-NVIDIA and that's on Android OS e.g. ARM, Qualcomm and Imagination Technologies. All 3 OpenGL GPU vendors dislikes NVIDIA's attempted fragmentation in the Android OS market.

You understand NVIDIA's OpenGL kit-bash extensions wasn't enough to negate December 3, 2014's Khronos Group's publically called for participants.

The majority of Khronos Group voted to dump the legacy OpenGL and select an API that was built for modern GPUs.

Mantle's value is to move the industry towards AMD's interest which is mostly the CPU side. If AMD has powerful CPU, these API changes would not be necessary.

NVidia's support for their kit bash doesn't matter. It weakened the mantle's position along side DX12 announcement.

"July 2014 to before February 19, 2015". You're timeline was not in order. Either way between December to March of 2014 AMD was struggling for support of the mantle project. Their involvement happened after or during those dates.

yeah agreed. the last good card they released in terms of all round performance (reliability, power to cost ratio, compatibility and general love for the product) was the 8800 back in the day. since then i have refused to go to nVidia, especially after working for apple for 10 years and knowing how shoddy their soldering is in the MBP's of days gone by (hence the shift to AMD)

I don't know why people are harping on Mantle. It's been handed over to the Khronos group to be implemented in Vulkan. I think that was a smart move by AMD. I hope to see Vulkan based games coming out in the next couple of years. Valve even talked about using Vulkan. Also, it helped push Microsoft to come up with DX 12 much faster. For anyone interested in what is going on with Vulkan here is what the Khronos group stated last month:

AMD is more friendly, but I bought nvidia cards last lineup.

Assholes or not, I buy the superior products.

Maxwell cards are just insane overclockers. All the 900 cards overclock really well. From my experience I've been able to get 1400-1500MHz on them easily. Hell my 980Ti gets 1380MHz @ stock with no OC...

@waahahah said:

Has no baring on the decisions made throughout mantle development. They explicitly stated time and time again to make it a public API.

Which is effectively fulfilled by Vulkan which is based from Mantle with improved cross-vendor compatibility.

@waahahah said:

GPUOpen being MIT doesn't matter since its still trademarked through them and that license is subject to change. Which is why intel/nvidia probably won't adopt it. It's the same with any real open source project, the owner of the source code can change the licensing on future versions of the software.

From http://download.intel.com/support/motherboards/server/sysmgmt/sb/sms_open_source_components_301111.pdf

Intel is currently using 3rd party components with MIT licensedfor Intel provided solutions.

GPUOpen is just a bundling for existing AMD's game software library and mostly targeting game developers. Hardware vendors can optimise GPUOpen for their hardware e.g. NVIDIA already optimised TressFX for their hardware.

Games such as Ashes of Singularity has equal source code access for AMD, Intel and NVIDIA.

@waahahah said:

the owner of the source code can change the licensing on future versions of the software.

Such arguments has no influence for the currentcopies of the software with attached MIT license. I can make my own future version of the software.

Game engines such Unity make use of Mono.net which is MIT licensed.

@waahahah said:

You're link does not state AMD will give the API to khronos, but they will give similar extensions... do even read the things you post?

"AMD, the red team will be supporting this open API with some high performance extensions that will offer almost similar performance to AMD’s upcoming API, Mantle."

Hence the word "like".

I'll it state gain. AMD's alternative plan for legacy OpenGL was to add Mantle likeextension.

The only person who is not reading is you.

@waahahah said:

You're link does not state AMD will give the API to khronos

History works against you. AMD's OpenGL vendor specific extension are no different to NVIDIA or Intel vendor specific OpenGL extensions.

AMD's Tiled Resource API was a vendor specific OpenGL extension before it was accepted and modified by Khronos i.e. API was renamed with ARB.

@waahahah said:

NVidia's support for their kit bash doesn't matter. It weakened the mantle's position along side DX12 announcement.

1. NVIDIA's OpenGL vendor specific kit-bash wasn't able to stop Khronos group calls for next-gen API. Also, AMD's OpenGL kit-bash is in the same boat as NVIDIA's OpenGL vendor specific kit-bash.

MS DirectX12 also weaken NVIDIA's OpenGL vendor specific kit-bash, hence the call for next gen API from Khronos group.

NVIDIA doesn't dominate OpenGL i.e. Qualcomm and ARM dominates OpenGL install base via Android OS. Qualcomm is a known anti-NVIDIA.

2. Your "NVidia's support for their kit bash doesn't matter" assertion doesn't address NVIDIA's OpenGL vendor specific kit-bash shipped PC game situation is worst than AMD Mantle.

3. AMD already knows Mantle's MS HLSL would not water in Khronos group i.e. it's either OpenGL's GLSL or OpenCL 2.1's SPIR-V.

4. Mantle on Windows has done it's job i.e. MS done their "assimilate and extend" tactics by releasing DX12 with AMD's CPU interest and it's very close to Mantle's programming model.

AMD’s Mantle was used as a disruptive force to change the industry.

AMD donated their efforts to the Khronos group and some of the relationship is obvious from the name.

Wrong. A view from Unity3D's Aras Pranckevičius http://aras-p.info/blog/2015/03/13/thoughts-on-explicit-graphics-apis/

Khronos Vulkan was announced (which is very much AMD Mantle, improved to make it cross-vendor).

Your post means squat when Mantle's API spec and it's programming model serves as the basis for Vulkan. It's the programmer that programs against APIs.

AMD isn't stopping anyone to write a Mantle wrapper AFAIK.

Your missing the point..... and making excuses now avoiding the fact, that their plans were early on was to keep it as an Trump card for themselves. Which it never happened... Since their actions in how Mantle was used and promoted when it was under control from AMD speaks how it was. The fact that parts of Mantle is apart of Vulkan again means squat since Vulkan is not Mantle nor is controlled by AMD.

The fact is that Mantle was only ever used by AMD products which means that its usage was proprietary by nature whatever the intent or plans were later on.

Again vulkan is a continuation of mantle, but no longer mantle, but this doesn't address the issue we are actually discussing. Let me remind you, AMD wanted to keep the API before making it open source.

From http://download.intel.com/support/motherboards/server/sysmgmt/sb/sms_open_source_components_301111.pdf

Intel is currently using 3rd party components with MIT licensedfor Intel provided solutions.

GPUOpen is just a bundling for existing AMD's game software library and mostly targeting game developers. Hardware vendors can optimise GPUOpen for their hardware e.g. NVIDIA already optimised TressFX for their hardware.

es such as Ashes of Singularity has equal source code access for AMD, Intel and NVIDIA.

Again your missing the point. NVIDIA optimizing for is because they have no choice. They will not promote or use it in any of their games. And again so long as AMD holds the licenses so Intel will likely get their own solution. Open source is still subject to change, it's not a permanent thing.

Such arguments has no influence for the currentcopies of the software with attached MIT license. I can make my own future version of the software.

Game engines such Unity make use of Mono.net which is MIT licensed.

Under a new branding. It means you will no longer be able to use new iterations of mantle and you'd have to get developer support. Also MIT license doesn't grant users the ability to change specification. Any changes will be extensions and probably difficult to get developer support. Mantle was NOT a collaboration project for the people that would need to use it. The licensing model doesn't really help them at all at this point, mantle needs to be standard for developers, if one day you can no longer use the standard, or make modifications to it, it no longer is standard and will less likely be supported by developers.

The MIT license really doesn't help an open standard, I can't make modifications to it because it will no longer be standard, and I'll have to fight AMD to get changes submitted as the standard.

Hence the word "like".

I'll it state gain. AMD's alternative plan for legacy OpenGL was to add Mantle likeextension.

The only person who is not reading is you.

No, you do not understand the argument or the issue we are talking about. Extensions are not mantle, mantle was intended to be proprietary from the beginning. Only making extensions for OpenGL reaffirm my argument that the Mantle was proprietary, was intended to be proprietary, but for public use.

History works against you. AMD's OpenGL vendor specific extension are no different to NVIDIA or Intel vendor specific OpenGL extensions.

AMD's Tiled Resource API was a vendor specific OpenGL extension before it was accepted and modified by Khronos i.e. API was renamed with ARB.

History is a bit irrelevant when AMD time and time again stated it was going to be a public API, provided by them.

1. NVIDIA's OpenGL vendor specific kit-bash wasn't able to stop Khronos group calls for next-gen API. Also, AMD's OpenGL kit-bash is in the same boat as NVIDIA's OpenGL vendor specific kit-bash.

MS DirectX12 also weaken NVIDIA's OpenGL vendor specific kit-bash, hence the call for next gen API from Khronos group.

NVIDIA doesn't dominate OpenGL i.e. Qualcomm and ARM dominates OpenGL install base via Android OS. Qualcomm is a known anti-NVIDIA.

2. Your "NVidia's support for their kit bash doesn't matter" assertion doesn't address NVIDIA's OpenGL vendor specific kit-bash shipped PC game situation is worst than AMD Mantle.

3. AMD already knows Mantle's MS HLSL would not water in Khronos group i.e. it's either OpenGL's GLSL or OpenCL 2.1's SPIR-V.

4. Mantle on Windows has done it's job i.e. MS done their "assimilate and extend" tactics by releasing DX12 with AMD's CPU interest and it's very close to Mantle's programming model.

You're missing the points entirely. Early 2014 AMD's plan was to provide the API and control the license model for it. It was going to be public which is a good thing but it was intended to be theirs. But they were struggling to get developer participation. I'll say this again... 12 games.

NVidia's kit bash served 1 purpose in my argument, from the perspective of developers, there were more standard alternatives.

DX12/Mantle are both probably a result of AMD/MS collaboration on xbox one. But who cares where it came from. Again from a developers perspective, there were more standard alternatives now.

AMD’s Mantle was used as a disruptive force to change the industry.

AMD donated their efforts to the Khronos group and some of the relationship is obvious from the name.

You are still missing the point. Mantle was only given to Khronos after it failed to get developer support to make it viable. AMD holding on to a dead technology was pointless so the only thing they could do was give it up to a 3rd party that intel/NVidia and developers would support.

It's no longer their say and in Khronos's hands, it can be a collaboration and intel/NVidia can probably make changes to it.