AMD CPU speculation... and expert conjecture

[palladin9479]

We ALL KNEW it
http://www.pcworld.com/article/2891672/amds-mantle-10-is-dead-long-live-directx.html

You should read articles fully before posting them.

Why this matters: If this is the end of Mantle 1.0 as we know it, it will have achieved what it needed to do—and that’s get Microsoft and OpenGL’s Khronos to finally move forward and faster on graphics APIs that have been holding back PC gaming.

AMD will keep it lurking around as insurance against standard stagnation. I'm impressed that they were able to push industry change so quickly against such massive entities without a fully blown standards war. Just the mere threat of that was enough to get things moving along. Hopefully we get stereoscopic rendering standardized so there isn't a need for a third party software / driver to do it.

 

[gamerk316]

Unsurprising news:

Back when it was announced, AMD were very bullish on the future of their Mantle API. They were confident it could help unleash the full power of gaming PCs, and they thought that developers would adopt it because of the advantages it offered.

But today, in a statement on their website, AMD explained that their plans for Mantle are changing rather drastically, and even recommended developers focus their efforts on other APIs, like DirectX 12.

http://www.pcgamesn.com/is-amd-walking-away-from-mantle

With Mantle being locked to AMD GCN hardware and Windows, with Apple rejecting it, with Valve rejecting it, with many announced games being canceled/delayed, with SDK in perennial beta stage, with DX12 being faster than Mantle, and with OGLNext coming, Mantle couldn't have any future...

So wait, back when everyone here was harping how everyone was going to use Mantle, who was one leading the charge saying it wouldn't take off?

Just like Pixel Shaders: The minute the vendors went off with their own implementations, they get mainlined in the DX API. That's what ALWAYS happens. The Vendors add vendor specific features, just implemented differently, and MSFT takes notice and puts them in the API.

 

[jed]

Unsurprising news:

Back when it was announced, AMD were very bullish on the future of their Mantle API. They were confident it could help unleash the full power of gaming PCs, and they thought that developers would adopt it because of the advantages it offered.

But today, in a statement on their website, AMD explained that their plans for Mantle are changing rather drastically, and even recommended developers focus their efforts on other APIs, like DirectX 12.

http://www.pcgamesn.com/is-amd-walking-away-from-mantle

With Mantle being locked to AMD GCN hardware and Windows, with Apple rejecting it, with Valve rejecting it, with many announced games being canceled/delayed, with SDK in perennial beta stage, with DX12 being faster than Mantle, and with OGLNext coming, Mantle couldn't have any future...

So wait, back when everyone here was harping how everyone was going to use Mantle, who was one leading the charge saying it wouldn't take off?

Just like Pixel Shaders: The minute the vendors went off with their own implementations, they get mainlined in the DX API. That's what ALWAYS happens. The Vendors add vendor specific features, just implemented differently, and MSFT takes notice and puts them in the API.

+1 you called it.

 

[8350rocks]

Unsurprising news:

Back when it was announced, AMD were very bullish on the future of their Mantle API. They were confident it could help unleash the full power of gaming PCs, and they thought that developers would adopt it because of the advantages it offered.

But today, in a statement on their website, AMD explained that their plans for Mantle are changing rather drastically, and even recommended developers focus their efforts on other APIs, like DirectX 12.

http://www.pcgamesn.com/is-amd-walking-away-from-mantle

With Mantle being locked to AMD GCN hardware and Windows, with Apple rejecting it, with Valve rejecting it, with many announced games being canceled/delayed, with SDK in perennial beta stage, with DX12 being faster than Mantle, and with OGLNext coming, Mantle couldn't have any future...

So wait, back when everyone here was harping how everyone was going to use Mantle, who was one leading the charge saying it wouldn't take off?

Just like Pixel Shaders: The minute the vendors went off with their own implementations, they get mainlined in the DX API. That's what ALWAYS happens. The Vendors add vendor specific features, just implemented differently, and MSFT takes notice and puts them in the API.

Well, to be fair...Vulkan is essentially 80% Mantle converted to run GLSL, and compatible with SPIR-V. They just tossed out the proprietary only bits and the stuff that was still being developed for later down the line.

In putting together SPIR-V, what Khronos has done is essentially extended Vulkan’s graphics constructs into the API, allowing SPIR-V to service both compute and graphics workloads. In the short term this is unlikely to make much of a difference for developers (who will be busy just learning the graphics side of Vulkan), but in the long run this would allow developers to more freely mix graphics and compute workloads, as the underlying runtime is all the same. This is also where Vulkan’s ability to extend its shading language from GLSL to other languages comes from, as SPIR’s flexibility is what allows multiple languages to all target SPIR.

That part is particularly relevant. It appears that the parts of Mantle that apply specifically to HSA APUs will be carried over directly into Vulkan, just as part of the cross platform API. Which, admittedly, could be used by Intel/Nvidia...however, the framework is there for the foundation. Which is a huge win for AMD in many respects.

 

[etayorius]

I love MANTLE and what it achieved, never used it though... I have no Radeon GPU to try it out.

I think we will see MANTLE 2.0 with hardware/software specific features, this way AMD can implement features not available on DX or OpenGLNxT which at some point it will force the other APIs to implement such features, in a way MANTLE has more say in the industry than what AMD ever had in Khronos or in DX with MS.

For now is mission accomplished.

I still can't forgive AMD and their 15 Plus Games in development with MANTLE to be released in 2014... jerks, they should never announce something they cannot accomplish.

 

[juanrga]

We ALL KNEW it
http://www.pcworld.com/article/2891672/amds-mantle-10-is-dead-long-live-directx.html

You should read articles fully before posting them.

Why this matters: If this is the end of Mantle 1.0 as we know it, it will have achieved what it needed to do—and that’s get Microsoft and OpenGL’s Khronos to finally move forward and faster on graphics APIs that have been holding back PC gaming.

AMD will keep it lurking around as insurance against standard stagnation. I'm impressed that they were able to push industry change so quickly against such massive entities without a fully blown standards war. Just the mere threat of that was enough to get things moving along. Hopefully we get stereoscopic rendering standardized so there isn't a need for a third party software / driver to do it.

When Mantle appeared, game developers such as John Carmack said that Mantle was providing the kind of performance already found in extensions to OGL

http://gearnuke.com/john-carmack-nvidia-opengl-extensions-amd-mantle-offer-similar-improvements-over-directx/

I think OGLNext (Vulkan) simply standardizes those old extensions. In fact, the people that will present OGLNext at GDC are VALVE/Nvidia people. Correct me if I am wrong but nobody from AMD will be at the OGLNext presentation talk.

Unsurprising news:

Back when it was announced, AMD were very bullish on the future of their Mantle API. They were confident it could help unleash the full power of gaming PCs, and they thought that developers would adopt it because of the advantages it offered.

But today, in a statement on their website, AMD explained that their plans for Mantle are changing rather drastically, and even recommended developers focus their efforts on other APIs, like DirectX 12.

http://www.pcgamesn.com/is-amd-walking-away-from-mantle

With Mantle being locked to AMD GCN hardware and Windows, with Apple rejecting it, with Valve rejecting it, with many announced games being canceled/delayed, with SDK in perennial beta stage, with DX12 being faster than Mantle, and with OGLNext coming, Mantle couldn't have any future...

So wait, back when everyone here was harping how everyone was going to use Mantle, who was one leading the charge saying it wouldn't take off?

We would differentiate "original Mantle" from "final Mantle".

Originally AMD promised more performance, multiplatform support (Apple and Linux were explicitly mentioned on talks), open API for everyone...

However, final Mantle is different. If AMD had maintained the original promises then the future of Mantle had been other. I started being critic of Mantle when the promises made by Hallock vanished in the air, and the DX12 pre-review just confirmed my suspicions.

 

[Embra]

I love MANTLE and what it achieved, never used it though... I have no Radeon GPU to try it out.

I think we will see MANTLE 2.0 with hardware/software specific features, this way AMD can implement features not available on DX or OpenGLNxT which at some point it will force the other APIs to implement such features, in a way MANTLE has more say in the industry than what AMD ever had in Khronos or in DX with MS.

For now is mission accomplished.

I still can't forgive AMD and their 15 Plus Games in development with MANTLE to be released in 2014... jerks, they should never announce something they cannot accomplish.

I am impressed with Mantle playing Dragon Age III. Stay high fps.

 

[blackkstar]

People are thinking that DX12 is going to replace Mantle. DX12 is not in a good position. It is Windows 10 only and Microsoft is releasing DX 11.3 to compete with DX12 and only to leave DX12 for developers who want lower level options.

In the future, you're going to see Mantle still be used by the types of developers who are more inclined to do technical stuff and push boundaries. DX11.3 will be used by developers who are opting for things like a AAA title with DLC cash grab.

I'm also assuming that Mantle will continue to serve as a beta testing ground for AMD and their partners to commit code to Vulkan. Companies like Oxide aren't going to abandon Mantle, they are doing great things with it and they have no governing body to try and petition to get changes made. I don't think Mantle will ever be a mainstream thing, but it will be the source of some cool tech demos and some good innovations to Vulkan.

Valve is rumored to be doing a lot of stuff with Vulkan, we'll have to see. But it sounds like, if FX 8350 is saying is true, that AMD just handed over large parts of Mantle to Khronos and Valve so Valve could take care of porting it to other platforms (like Linux/SteamOS) and they wouldn't have to deal with that. Mantle is designed to push hardware in ways previous APIs couldn't. AMD simply doesn't have the resources (even though they say they do) to waste maintaining an API on Linux, Windows, Android, etc. like Vulkan will support. They are supposed to be giving a good talk on Vulkan today, actually. There's also something with Oxide and AMD with Mantle that's supposed to be announced.

It's too early to completely write off Mantle. The developers who would rather create "cinematic experiences" and "some safe-market game with loads of DLC" will all abandon Mantle for DX12 and Vulkan. But the developers like Oxide, who like to push tech, are going to stay with Mantle.

The biggest winner out of all of this seems to be Vulkan. It's going to be far more open than Mantle or DX12, and it's not going to be limited to a specific OS or specific hardware. I'm assuming most people are going to take Vulkan on whatever OS they want (and this includes Windows 7!) as opposed to jumping into Windows 10.

Vulkan is going to be interesting for Nvidia. It relies heavily on OpenCL and Nvidia has traditionally "not optimized" (or actively harmed) OpenCL performance to encourage developers to stay with CUDA.

http://www.phoronix.com/scan.php?page=article&item=khronos-vulcan-spirv&num=1

The most important thing though, is that things are changing. We're going to get a lot of new, exciting things at a time when graphics have grown stale and not improving so much, as evidenced by 3 year old graphic cards being more than enough and people being reduced to arguing about performance on 4k and 8k resolutions because the common gaming resolutions are completely covered by mid range cards.

Also, I called this long ago, and I'm going to call it now. The future of graphic APIs will be Nvidia pushing Windows 10 and DX12 against the entire world. Nvidia has nothing to gain from Mantle and it looks like all their sabotaging OpenCL performance over all these years is going to really hurt their Vulkan performance since Vulkan relies so heavily on OpenCL.

http://developer.amd.com/community/events/amd-gdc-2015/
It looks like the big Oxide Games tech demo that I've seen kicked around is on March 5th.

 

[juanrga]

People are thinking that DX12 is going to replace Mantle. DX12 is not in a good position. It is Windows 10 only and Microsoft is releasing DX 11.3 to compete with DX12 and only to leave DX12 for developers who want lower level options.

In the future, you're going to see Mantle still be used by the types of developers who are more inclined to do technical stuff and push boundaries. DX11.3 will be used by developers who are opting for things like a AAA title with DLC cash grab.

I'm also assuming that Mantle will continue to serve as a beta testing ground for AMD and their partners to commit code to Vulkan. Companies like Oxide aren't going to abandon Mantle, they are doing great things with it and they have no governing body to try and petition to get changes made. I don't think Mantle will ever be a mainstream thing, but it will be the source of some cool tech demos and some good innovations to Vulkan.

Valve is rumored to be doing a lot of stuff with Vulkan, we'll have to see. But it sounds like, if FX 8350 is saying is true, that AMD just handed over large parts of Mantle to Khronos and Valve so Valve could take care of porting it to other platforms (like Linux/SteamOS) and they wouldn't have to deal with that. Mantle is designed to push hardware in ways previous APIs couldn't. AMD simply doesn't have the resources (even though they say they do) to waste maintaining an API on Linux, Windows, Android, etc. like Vulkan will support. They are supposed to be giving a good talk on Vulkan today, actually. There's also something with Oxide and AMD with Mantle that's supposed to be announced.

It's too early to completely write off Mantle. The developers who would rather create "cinematic experiences" and "some safe-market game with loads of DLC" will all abandon Mantle for DX12 and Vulkan. But the developers like Oxide, who like to push tech, are going to stay with Mantle.

The biggest winner out of all of this seems to be Vulkan. It's going to be far more open than Mantle or DX12, and it's not going to be limited to a specific OS or specific hardware. I'm assuming most people are going to take Vulkan on whatever OS they want (and this includes Windows 7!) as opposed to jumping into Windows 10.

Vulkan is going to be interesting for Nvidia. It relies heavily on OpenCL and Nvidia has traditionally "not optimized" (or actively harmed) OpenCL performance to encourage developers to stay with CUDA.

http://www.phoronix.com/scan.php?page=article&item=khronos-vulcan-spirv&num=1

The most important thing though, is that things are changing. We're going to get a lot of new, exciting things at a time when graphics have grown stale and not improving so much, as evidenced by 3 year old graphic cards being more than enough and people being reduced to arguing about performance on 4k and 8k resolutions because the common gaming resolutions are completely covered by mid range cards.

Also, I called this long ago, and I'm going to call it now. The future of graphic APIs will be Nvidia pushing Windows 10 and DX12 against the entire world. Nvidia has nothing to gain from Mantle and it looks like all their sabotaging OpenCL performance over all these years is going to really hurt their Vulkan performance since Vulkan relies so heavily on OpenCL.

http://developer.amd.com/community/events/amd-gdc-2015/
It looks like the big Oxide Games tech demo that I've seen kicked around is on March 5th.

About Mantle, the article given by jdwii contains a very relevant quote from a game developer:

Frankly, the only reason I would support Mantle is if AMD paid me to.

The relations between Vulkan and Nvidia was addressed only a pair of posts above.

 

[etayorius]

People are thinking that DX12 is going to replace Mantle. DX12 is not in a good position. It is Windows 10 only and Microsoft is releasing DX 11.3 to compete with DX12 and only to leave DX12 for developers who want lower level options.

In the future, you're going to see Mantle still be used by the types of developers who are more inclined to do technical stuff and push boundaries. DX11.3 will be used by developers who are opting for things like a AAA title with DLC cash grab.

I'm also assuming that Mantle will continue to serve as a beta testing ground for AMD and their partners to commit code to Vulkan. Companies like Oxide aren't going to abandon Mantle, they are doing great things with it and they have no governing body to try and petition to get changes made. I don't think Mantle will ever be a mainstream thing, but it will be the source of some cool tech demos and some good innovations to Vulkan.

Valve is rumored to be doing a lot of stuff with Vulkan, we'll have to see. But it sounds like, if FX 8350 is saying is true, that AMD just handed over large parts of Mantle to Khronos and Valve so Valve could take care of porting it to other platforms (like Linux/SteamOS) and they wouldn't have to deal with that. Mantle is designed to push hardware in ways previous APIs couldn't. AMD simply doesn't have the resources (even though they say they do) to waste maintaining an API on Linux, Windows, Android, etc. like Vulkan will support. They are supposed to be giving a good talk on Vulkan today, actually. There's also something with Oxide and AMD with Mantle that's supposed to be announced.

It's too early to completely write off Mantle. The developers who would rather create "cinematic experiences" and "some safe-market game with loads of DLC" will all abandon Mantle for DX12 and Vulkan. But the developers like Oxide, who like to push tech, are going to stay with Mantle.

The biggest winner out of all of this seems to be Vulkan. It's going to be far more open than Mantle or DX12, and it's not going to be limited to a specific OS or specific hardware. I'm assuming most people are going to take Vulkan on whatever OS they want (and this includes Windows 7!) as opposed to jumping into Windows 10.

Vulkan is going to be interesting for Nvidia. It relies heavily on OpenCL and Nvidia has traditionally "not optimized" (or actively harmed) OpenCL performance to encourage developers to stay with CUDA.

http://www.phoronix.com/scan.php?page=article&item=khronos-vulcan-spirv&num=1

The most important thing though, is that things are changing. We're going to get a lot of new, exciting things at a time when graphics have grown stale and not improving so much, as evidenced by 3 year old graphic cards being more than enough and people being reduced to arguing about performance on 4k and 8k resolutions because the common gaming resolutions are completely covered by mid range cards.

Also, I called this long ago, and I'm going to call it now. The future of graphic APIs will be Nvidia pushing Windows 10 and DX12 against the entire world. Nvidia has nothing to gain from Mantle and it looks like all their sabotaging OpenCL performance over all these years is going to really hurt their Vulkan performance since Vulkan relies so heavily on OpenCL.

http://developer.amd.com/community/events/amd-gdc-2015/
It looks like the big Oxide Games tech demo that I've seen kicked around is on March 5th.

Someone care to explain what the hell is Vulkan? been seeing people commenting about it but i still have no idea what it is or what it does, is this some sort of API?

 

[-Fran-]

Someone care to explain what the hell is Vulkan? been seeing people commenting about it but i still have no idea what it is or what it does, is this some sort of API?

http://www.phoronix.com/scan.php?page=news_item&px=Khronos-Vulkan-Graphics-API

Basically, Vulkan is the code name for OGLN.

Cheers!

EDIT: MARKETING SLIIIIIIIIIIIIIIIIIIIDES! http://www.phoronix.com/scan.php?page=article&item=khronos-vulcan-spirv&num=1

 

[gamerk316]

A bit off topic, but here's a point that puts this all in perspective: Windows XP market share is inching up.

 

[-Fran-]

A bit off topic, but here's a point that puts this all in perspective: Windows XP market share is inching up.

That's a bit hard to believe... Do you have any source?

Cheers!

 

[Cazalan]

Vulkan is a fork of Mantle, therefore Mantle lives on.


http://forums.anandtech.com/showpost.php?p=37218504&postcount=56

"If you compare the function names you'll see that the Vulkan functions are named identical to Mantle, and the only difference is the prefix."

 

[logainofhades]

A bit off topic, but here's a point that puts this all in perspective: Windows XP market share is inching up.

Not surprised really, with the flop that was Win 8/8.1.

 

[gamerk316]

A bit off topic, but here's a point that puts this all in perspective: Windows XP market share is inching up.

That's a bit hard to believe... Do you have any source?

Cheers!

Take the source as you will:

http://www.fudzilla.com/news/37162-windows-xp-will-not-die

 

[etayorius]

Vulkan is a fork of Mantle, therefore Mantle lives on.


http://forums.anandtech.com/showpost.php?p=37218504&postcount=56

"If you compare the function names you'll see that the Vulkan functions are named identical to Mantle, and the only difference is the prefix."

VULKAN with MANTLE at the core, seems legit and quite the coincidence.

I hope Vulkan becomes the standard and DX12 fades into oblivion.

 

[-Fran-]

Take the source as you will:

http://www.fudzilla.com/news/37162-windows-xp-will-not-die

That's a weird occurrence indeed. Do you remember if such thing happened with Win98/2K during the transition to WinXP?

I know WinXP is rock solid at this point, but I wouldn't go back to it from Win7. I mean, unless I have specific needs, I don't see why I should go back to it. So, XP re-gaining market share, is bizarre to say the least.

Cheers!

 

[jdwii]

XP usage increase is probably not in the US, anyways i posted that article really late last night about Mantle being dead. Guessing it will become more of a AMD thing and will probably be used when Amd pays people to use it. I see no reason at all for Intel or Nvidia(majority) to use it now.

http://www.maximumpc.com/amd_developers_focus_your_efforts_directx_12_not_mantle_2015

"AMD told developers that if they're interested in Mantle 1.0's functionality, they should focus their attention on DirectX 12 or GLnext" Pretty sure that means dead when even Amd claims to use other stuff.

 

[8350rocks]

Vulkan is a fork of Mantle, therefore Mantle lives on.


http://forums.anandtech.com/showpost.php?p=37218504&postcount=56

"If you compare the function names you'll see that the Vulkan functions are named identical to Mantle, and the only difference is the prefix."

See, juanrga, AMD delivered, they just did it in a round about way. Mantle is THE open source API you were looking for on Linux.

Khronos just dropped support for HLSL in favor of GLSL until they come up with something more cross platform friendly...(there is discussion about being able to take C++ code straight into SPIR-V and make that work...we shall see though...)

 

[8350rocks]

XP usage increase is probably not in the US, anyways i posted that article really late last night about Mantle being dead. Guessing it will become more of a AMD thing and will probably be used when Amd pays people to use it. I see no reason at all for Intel or Nvidia(majority) to use it now.

http://www.maximumpc.com/amd_developers_focus_your_efforts_directx_12_not_mantle_2015

"AMD told developers that if they're interested in Mantle 1.0's functionality, they should focus their attention on DirectX 12 or GLnext" Pretty sure that means dead when even Amd claims to use other stuff.

Negative, Vulkan is basically mantle, as I said above...AMD just handed it over to somebody who can do more with it...

 

[juanrga]

See, juanrga, AMD delivered, they just did it in a round about way. Mantle is THE open source API you were looking for on Linux.

Negative, Vulkan is basically mantle, as I said above...AMD just handed it over to somebody who can do more with it...

Vulkan must be the API I was expecting, but it is not Mantle:

we need to be clear that Vulkan is not Mantle, Mantle was used to bootstrap the process and speed its development

This quote from Kronos group president (Nvidia vicepresident) is also relevant:

Any hardware capable of supporting OpenGL ES 3.1 will be capable of supporting Vulkan.

Any modern hardware would be Vulkan compatible, which would help popularization. Here a partial list of compatible hardware

http://en.wikipedia.org/wiki/OpenGL_ES#OpenGL_ES_3.1_2

What Does The Future Hold For AMD's Mantle?

http://www.tomshardware.com/news/amd-mantle-directx-12-vulkan,28676.html

 

[-Fran-]

See, juanrga, AMD delivered, they just did it in a round about way. Mantle is THE open source API you were looking for on Linux.

Negative, Vulkan is basically mantle, as I said above...AMD just handed it over to somebody who can do more with it...

Vulkan must be the API I was expecting, but it is not Mantle:

we need to be clear that Vulkan is not Mantle, Mantle was used to bootstrap the process and speed its development

This quote from Kronos group president (Nvidia vicepresident) is also relevant:

Any hardware capable of supporting OpenGL ES 3.1 will be capable of supporting Vulkan.

Any modern hardware would be Vulkan compatible, which would help extension.

I like this quote: "Vulkan is not MANTLE just like OpenGL is not Glide".

Cheers!

EDIT: MARKETING SLIIIIIIDES for VR: http://www.tomshardware.com/news/amd-liquidvr-virtual-reality,28682.html

That's what will push compute and HSA forward IMO.

 

[jdwii]

Just because a Nexus 7 does the same thing as a Ipad doesn't make them the same product. 4+4=8 and 2+6= 8 same outcome however different.

Anyways its still a win for Amd and everyone else. Nvidia had to do nothing really to even get it.

 

[jdwii]

Also completly irreverent to this thread but Nvidia's show was HIGHLY disappointing. I was expecting to see a 980 Ti or a new titan class video card instead we got trolled into a product no one in this forum will even want.