The AMD Mantle Thread

Page 7 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Status
Not open for further replies.
Aug 11, 2008
10,451
642
126
So suddenly you have a problem with DirectX on your nvidia card? For you nothing changes. You still get your directx. Nobody will make mantle only games.

But, apparently, now you have a problem with directx being "crappy". Whatever, if you say so. Sounds like sour grapes to me. Nothing changes for you. AMD will get a couple of games that benefits them, kinda like (and this isn't an apples to apples comparison) how nvidia users benefit from physx. AMD gets a few games a year that benefits them. So what?

Again, sounds like nothing more than sour grapes. Nothing changes for you. You continue using DirectX. Although, hilariously, you seem to have problems with DX being "crappy" now. There is also no fragmentation. Maybe you should read up on what the 3d rendering situation was like back in 1996 where some people could use 3d acceleration and some could not. That is not the case now. Again, you still have your 3d accelerated DirectX. Nothing has and nothing will change for you - nvidia still excels in doing their things with DX11 games.

What you are failing to acknowledge is that there is a limited amount of resources a developer is willing to devote to porting a game to PC. Sure directX will continue, unless Valve decides to make SteamOS specific games. But if the resources for porting a game to a PC now must be divided among SteamOS, Mantle, and DirectX, I dont see how that will not reduce the quality of the DirectX version.
 

Carfax83

Diamond Member
Nov 1, 2010
6,841
1,536
136
Intel already does by using/creating cpu specific extensions.

Yeah, but Intel and AMD use the same basic x86-64 architecture, and both have a cross licensing agreement, so AMD usually implement those extensions themselves in CPUs.

NVidia and AMD GPUs on the other hand, are completely different architecture wise.
 

Carfax83

Diamond Member
Nov 1, 2010
6,841
1,536
136
But you also forget that a "low level" API from my understanding means the exposure of that particular graphics architecture (for someone to have a "peek" of that particular design) i.e. developers can gain direct access to the hardware and even within the same family of GPUs, the code will be different as they have different configurations of different logic parts.

I don't think Mantle is as low level as you say. Judging by this pic:



The mantle API sits between the 3D program itself, and the mantle driver, much like DirectX but just a lot more efficient since it doesn't have to worry about preserving backward compatibility.

I don't know, the more I read about Mantle, the more I think it may not be so bad after all. I'm pretty sure the code won't need to be re-written or changed for every GPU within the same families. That would make it far more expensive to implement, no developer would do it.

If NVidia could use Mantle for their own GPUs, by re-writing the code, then it could be a win win situation.
 
Last edited:

OCGuy

Lifer
Jul 12, 2000
27,227
36
91
Is Also the vibe I'm getting from the developers that are commenting about Mantle is that much of the benefit comes from having control over memory which is often a neglected aspect of these types of discussions.

Maxwell is going to give the GPU access to system memory, so there is a lot coming up on that front.
 

bononos

Diamond Member
Aug 21, 2011
3,894
162
106
Would AMD slot in a 'cripple Intel' check inside Mantle heh heh like Intel did with their own compilers.
 

boxleitnerb

Platinum Member
Nov 1, 2011
2,601
2
81
http://www.pcgameshardware.de/Grafi...D-Low-Level-API-Mantle-Battlefield-4-1090085/

PCGH asked Raja Koduri about Mantle and its "openess". He replied that AMD doesn't see Mantle as an open standard like OpenCL or OpenGL. He also tried to brush aside comparisons with Glide but then stated: If a competitor were to approach AMD to make their own backend and drivers for Mantle, AMD would not dismiss them right away.

That doesn't sound very open to me. In the letter of the word it may be "open", but in the spirit, this thing is clearly proprietary since AMD has the final say and there is no independent group involved in its inception and ongoing development. This leaves a very sour taste in my mouth, no matter the possible advantages.
 

wand3r3r

Diamond Member
May 16, 2008
3,180
0
0

blastingcap

Diamond Member
Sep 16, 2010
6,654
5
76
For the past several years, I reluctantly bought AMD cards because they undercut NV on price/perf. All else equal I prefer NV. But if AMD somehow pulls this off (requires massive developer support... and perhaps talking to Valve and getting it to work on Linux) and NV GPUs do not work as well with it if at all, I don't see why any PC gamers would opt for NV assuming equal price. PhysX? That said, I do not plan to buy any more 28nm GPUs, so I have some time to wait-and-see.
 
Last edited:

boxleitnerb

Platinum Member
Nov 1, 2011
2,601
2
81
What's your thought on physx?

I've always seen PhysX as a temporary solution until the new consoles and until an open usable solution were established, respectively. Because I knew, only then could an open solution really take off since the developers are hesitant about programming stuff exclusively (and expensively) for the PC alone. PhysX is a middleware - the comparison between Mantle and CUDA is much more logical. And CUDA was developed when OpenCL was not even ratified in its first version. Mantle on the other hand comes at a time when there are viable alternatives. If not DirectX, then OpenGL. Do we need a third one?

Besides, are you honestly comparing a couple of effects in 1-2 games/year with a whole new API that has the potential to split the gaming market in the middle? Well, I'm not.
 

SiliconWars

Platinum Member
Dec 29, 2012
2,346
0
0
https://twitter.com/AMDRadeon/status/383349757428506624

This is the second AMD source to point people in the direction of Ryan's article = strong possibility of it being true. IF Mantle is the Xbox's low level API, it's probably game over or a seismic shift at minimum. IF both the Xbox and PS4 have it, it's definitely all over, no way can Nvidia withstand that kind of loss.
 

sontin

Diamond Member
Sep 12, 2011
3,273
149
106
What's your thought on physx?

PhysX is a physics middleware running on different CPU architectures (x86, ARM, PowerPC etc.) and Cuda GPUs down to the first Cuda GPU G80.

Mantle is a graphics API only running on GCN GPUs.

Hm. :hmm:
 
Feb 19, 2009
10,457
10
76
https://twitter.com/AMDRadeon/status/383349757428506624

This is the second AMD source to point people in the direction of Ryan's article = strong possibility of it being true. IF Mantle is the Xbox's low level API, it's probably game over or a seismic shift at minimum. IF both the Xbox and PS4 have it, it's definitely all over, no way can Nvidia withstand that kind of loss.

It all depends on what kind of perf gains Mantle brings over DX, if its around 30-50%, NV can still compete by going brute force big dies, since they have Quadro and Tesla products to cover the costs of making large dies and then some.

If its >50%... lets hope Maxwell is all its cracked up to be.
 

raghu78

Diamond Member
Aug 23, 2012
4,093
1,475
136
Yeah, but Intel and AMD use the same basic x86-64 architecture, and both have a cross licensing agreement, so AMD usually implement those extensions themselves in CPUs.

NVidia and AMD GPUs on the other hand, are completely different architecture wise.

Intel and AMD implement the same x86-64 bit ISA (instruction set architecture). their respective CPU microarchitectures Ivybridge / Haswell and Piledriver / Steamroller are vastly different. Its similar to AMD and NV supporting Microsoft Directx HLSL and OpenGL but their underlying architectures GCN and Kepler are vastly different.
 
Last edited:

Saylick

Diamond Member
Sep 10, 2012
3,385
7,151
136
It all depends on what kind of perf gains Mantle brings over DX, if its around 30-50%, NV can still compete by going brute force big dies, since they have Quadro and Tesla products to cover the costs of making large dies and then some.

If its >50%... lets hope Maxwell is all its cracked up to be.

A guaranteed +30% to +50% to GCN cards would destroy any line-up nVidia offers imo. That kind of efficiency gain would allow AMD to release small dies which would be able to compete against nVidia's large dies, a la AMD 4xxx/5xxx vs nVidia 2xx/4xx AND also enable them to build slightly smaller dies (perhaps ~425 mm^2) which would enable AMD to grab the performance crown generation after generation.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
It all depends on what kind of perf gains Mantle brings over DX, if its around 30-50%, NV can still compete by going brute force big dies, since they have Quadro and Tesla products to cover the costs of making large dies and then some.

If its >50%... lets hope Maxwell is all its cracked up to be.

How do you even come up with those numbers?

Do you really think there is even 30%-50% to be gained? And not more in the area of say 5% or less?

This is a classic hype.
 
Feb 19, 2009
10,457
10
76
How do you even come up with those numbers?

Do you really think there is even 30%-50% to be gained? And not more in the area of say 5% or less?

The devs people love to quote, often throw around DX overhead on hardware as in the 2.5x range, ie. needing 2.5x the power to achieve the same as a low level API would.

I am being very conservative, *IF* Mantle is indeed an effective low level API as they claimed, 30-50% is the low ball estimate.

* It's a big IF since we haven't got any concrete results yet. It could well be "meh".. but "meh" would not excite DICE, nor would it be brewing for 2 years.
 

boxleitnerb

Platinum Member
Nov 1, 2011
2,601
2
81
I guess predominantly it will impact CPU usage. How large the effect on GPUs will be, remains to be seen.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
The devs people love to quote, often throw around DX overhead on hardware as in the 2.5x range, ie. needing 2.5x the power to achieve the same as a low level API would.

I am being very conservative, if Mantle is indeed an effective low level API as they claimed, 30-50% is the low ball estimate.

You are in for a big dissapointment then.
 
Feb 19, 2009
10,457
10
76
Dont try and twist it. Its you claiming the overhyped numbers. Good luck with that, you will need it.

No sir. Not claiming, merely speculating. I don't know how effective Mantle is, I am just using the common agreed DX overhead on hardware, and then being very conservative, I am guessing it could be 30-50%. Its as good as anyone else's guesses at this point.

Now, why exactly are you so certain it won't be any good?
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
No sir. Not claiming, merely speculating. I don't know how effective Mantle is, I am just using the common agreed DX overhead on hardware, and then being very conservative, I am guessing it could be 30-50%. Its as good as anyone else's guesses at this point.

Now, why exactly are you so certain it won't be any good?

DX10-11 would essentially have to be the biggest turd alive for that to be true. Not to mention how DX9 would be. Or how big an epic flop OpenGL would have to be. All fixed by magic fairies to gain minimum 30-50% on a new API.

There is a reason why AMD didnt show any real gain numbers. Its nothing great to write home about.

Will it be better than DX? Most likely. But its not gonna be near those hype numbers.
 
Feb 19, 2009
10,457
10
76
DX10-11 would essentially have to be the biggest turd alive for that to be true. Not to mention how DX9 would be. Or how big an epic flop OpenGL would have to be. All fixed by magic fairies to gain minimum 30-50% on a new API.

There is a reason why AMD didnt show any real gain numbers. Just likely why no 290X specs. Its nothing great to write home about.

So I take it you are a game developer who understands all the nuisance of DX programming? You realize big developers are the ones making the claims, that DX actually does have a massive overhead compared to coding to the metal as on console APIs??

It's exactly what Raja said, coding to the metal for PC is pretty much giving you 2018 hardware today. Now, obviously he works for AMD so he will hype it to the max. So I am more conservative, instead of massive 2.5x gains, I'll settle for 30-50%
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
So I take it you are a game developer who understands all the nuisance of DX programming? You realize big developers are the ones making the claims, that DX actually does have a massive overhead compared to coding to the metal as on console APIs??

It's exactly what Raja said, coding to the metal for PC is pretty much giving you 2018 hardware today. Now, obviously he works for AMD so he will hype it to the max. So I am more conservative, instead of massive 2.5x gains, I'll settle for 30-50%

We should heed Mr. "No AMD APUs for next-gen consoles" yo. Massive credentials right there.
 
Status
Not open for further replies.
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |