• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

AMDs Future of Compute: CARRIZO APU, 5 Samsung UHD FreeSync displays + more

DonMigs85

Member
As much as I want to give AMD business, their single-threaded performance still sucks compared to Intel's and they're still stuck on 28nm when Intel's been on 22nm for a while and will pretty soon move to 14nm.
Plus you have to buy really high-speed DDR3 to get the most out of the IGP which negates some of the cost savings here. They probably should have adopted DDR4 instead.
 

Locuza

Member
I'm looking forward to Carrizos GPU performance, since Delta Color Compression should help a lot.
But the next real chance will come 2016 with "14/16" nm + FinFETs and Zen/K12.
 

DonMigs85

Member
I'm looking forward to Carrizos GPU performance, since Delta Color Compression should help a lot.
But the next real chance will come 2016 with "14/16" nm + FinFETs and Zen/K12.

Hopefully by then they'll have abandoned the vile Bulldozer architecture.
 

Datschge

Member
You could show an old, blind man, where he can find the sentence, stating DICE pitched the idea first to Nvidia.
Please don't leave the thread now.
It doesn't, Johan Andersson only states that they (as in DICE) started thinking about now 6 years ago ("Vor ungefähr fünf Jahren fingen wir an darüber nachzudenken.") and talked to different companies including Intel and Nvidia, without specifying any order ("Wir haben mit verschiedenen Firmen gesprochen, inklusive Intel und Nvidia.").

In the next question he states that while he was not the only one doing the groundwork, he was the one pushing the most for it ("Q: Also haben Sie quasi den Grundstein für Mantle gelegt? A: Natürlich nicht ich allein, aber ich war wahrscheinlich derjenige, der sich am meisten ins Zeug gelegt hat."). Then continues how they had to build a new render backend at DICE, and AMD had to assemble a driver team to develop the alternate 3D interface, for which he respect AMD as none of the other companies did that ("Mantle bedeutete für uns bei DICE, ein komplett unterschiedliches Render-Backend zu kreieren. Für AMD hieß es ein Treiberteam aufzubauen, welches seine Ressourcen und Zeit in die Entwicklung einer alternativen 3D-Schnittstelle steckt. Das alles sind große Schritte und ich muss AMD Respekt zollen, dass sie meine Vorschläge umgesetzt haben. Denn keiner der anderen Hersteller hat das getan.").

In the same interview he goes on that the interface is hardware agnostic ("Mantle ist keine exklusive, konsolenartige AMD-Schnittstelle für GCN-Grafikchips. Mit Mantle ist es möglich, Feinheiten der GPUs auszunutzen. Zu spezifische Sachen lassen sich dagegen über Extensions realisieren. Es ist für andere Hersteller möglich, Mantle zu unterstützen."), that he wants it on mobile, or rather "all", systems ("Q: Ist Mantle künftig auch auf Smartphones und Tablets denkbar? A: Absolut. Ich hätte Mantle gern überall."), that he wants to see it under different OSes ("Gern würde ich Mantle auch unter Linux und Mac OS sehen. Es ist jedenfalls denkbar. Das ist meine Vision: Mantle auf unterschiedlichen Plattformen.").

It's up to you if you think if he's talking out of his ass, he and DICE actually had no input and Mantle is a big PR scheming by AMD to claim openness while so far only doing the opposite only with Mantle. While that willfully doing non-open stuff is the thing Nvidia fans appear to applaud whenever they come up with something new. *shrugs*
 

Locuza

Member
It's up to you if you think if he's talking out of his ass, he and DICE actually had no input and Mantle is a big PR scheming by AMD to claim openness while so far only doing the opposite only with Mantle. While that willfully doing non-open stuff is the thing Nvidia fans appear to applaud whenever they come up with something new. *shrugs*
First of all, thanks for the translation, but my native language is german.
Now for the section I quoted, where did i claim such things?

I never wrote that Johan is talking out of his ass, neither that DICE had no input (I even said in the text you quoted, that DICE helped a lot) and I also never said that Mantle is just a big PR monument.

What i'm trying to say is that the whole thing about being open/cooperative is inaccurate and came to early.
There exist many silly things, like Nvidia could support Mantle if they want to but Huddy thinks their pride will stop them from doing so.
Which is of course very sad (having a functional brain).

Looking at Nvidia, I hate the anti FreeSync politics , I didn't applaud about GameWorks and many other things.
I shaked the head about one Interview with Nvidia representants because they also tried to manipulate the view of certain things.
 
I don't hate AMD at all. In fact, I have an Athlon FX processor and an ATI Radeon 9700 displayed in my glass cabinet.

What I really hate is marketing something as "open" which isn't open in the least, so I haven't been too happy about them since Mantle. And I guess I'm also highly iritated at some AMD fanboys.

Sure:

From what I understand, Mantle is invitation only at this point until it comes out of beta, AMD themselves said that it'll be released to the public when development finishes. At that point, Mantle 2.0 would be developed.
 

Qassim

Member
The whole idea that the first time the public would see an open standard API is over a year after it was created with an SDK release is, frankly, ludicrous.

Yep, and when you look at it from the perspective of NVIDIA and Intel - why would they support that? Why would they then support something that has been kept away from them for a year whilst their competitor has utilised it for whatever advantage it offered?

If AMD were ever serious about this being a collaborative free ('not as in beer') and open 'standard' or project, they'd have done this at the very start. Not a year later, lol.

From what I understand, Mantle is invitation only at this point until it comes out of beta, AMD themselves said that it'll be released to the public when development finishes. At that point, Mantle 2.0 would be developed.

Which is exactly how you don't run a supposedly open project such as this, which it is being sold as.
 

wachie

Member
You know, you could at least read the entire post you're quoting.


I made a post on how Mantle is not in any way, shape or form an open standard in an earlier thread.

The whole idea that the first time the public would see an open standard API is over a year after it was created with an SDK release is, frankly, ludicrous.
Maybe you should also read my post for starters? Why don't you point out such ludocrities (some even worse) for Nvidia?
 

blu

Wants the largest console games publisher to avoid Nintendo's platforms.
Because Nvidia isn't pretending to make a 'open source" API
Correction: 'open standard', not 'open source'.

NV support OGL to the fullest of their abilities, even though it's a 'design by committee' standard. AMD are trying to do something with Mantle which is not really clear - supposedly it has technical merits beyond DX11/current GL, but those quickly blur where it comes to DX12/GL next. And it surely is not an open standard, not by a long shot.
 
From what I understand, Mantle is invitation only at this point until it comes out of beta, AMD themselves said that it'll be released to the public when development finishes. At that point, Mantle 2.0 would be developed.

I think they will kill mantle once Dx12 and the new OpenGL are released.
Probably more to put some heat under microsoft and khronos group, but then
again i could be totally wrong.
 

Durante

Member
Correction: 'open standard', not 'open source'.

NV support OGL to the fullest of their abilities, even though it's a 'design by committee' standard. AMD are trying to do something with Mantle which is not really clear - supposedly it has technical merits beyond DX11/current GL, but those quickly blur where it comes to DX12/GL next. And it surely is not an open standard, not by a long shot.
I think it's very clear: they are trying to leverage their console position in order to gain an advantage in the PC space. Which is perfectly fine.

Were they completely lost me is in pushing the marketing idea that it is an open standard (which, as you note, it really isn't by a long shot), and that Intel and NV are somehow to blame for not signing their own death warrant.
 

Kezen

Banned
AMD will never make Mantle a "standard" without Nvidia and Intel's support.
http://www.techradar.com/news/computing/pc/amd-on-mantle-we-want-our-gaming-api-to-become-the-industry-standard-1218560

AMD told TechRadar that it'd be willing to make Mantle, or an API based on it, available across the industry. Even, the company said, if it means Mantle is adopted by competitors like Nvidia.
"Mantle for now is straight up in a closed ecosystem, a closed beta, which you have to do in a complicated project like this to get it off the ground. It's us and a few key game developers," Robert Hallock, technical communications, AMD Graphics & Gaming, told us in a recent interview.

"After that phase is done, we do hope that Mantle becomes an industry standard. We'll be releasing a public SDK later this year, and hope that others adopt it. If they don't adopt it itself, then we hope they adopt APIs similar to it that become an industry standard for PC gaming."

Good luck with that AMD.
 
Hopefully by then they'll have abandoned the vile Bulldozer architecture.

I never got why they called it Bulldozer. Or any of their other silly construction codenames like Piledriver, Steamroller, or Excavator. Because when I think of fast things Bulldozer is the first thing that comes to mind, right?
 
I never got why they called it Bulldozer. Or any of their other silly construction codenames like Piledriver, Steamroller, or Excavator. Because when I think of fast things Bulldozer is the first thing that comes to mind, right?

Construction codenames?

Man, I always just thought AMD were perverts like the rest of us.
 

Datschge

Member
Good luck with that AMD.
At the current rate OpenGL Next may well contain Mantle 2.0 ideally making Mantle itself completely obsolete. The advantages for all parties are obvious: Mantle beta developers are prepared, Khronos Group got a free beta testing round for a new API, and AMD removes itself from the awkward chicken egg problem of rivals naturally having no interest in implementing its proprietary API, and AMD itself having not enough financial, marketing and man power to drive it all alone.

(For the record, the news that for OpenGL Next AMD has given Khronos Group unfettered access to Mantle to take as many pages as it wants out of the Mantle playbook, and AMD imposing no restrictions as well as no licensing fees, is already over three months old by now.)
 
Top Bottom