Nvidia or ATI gpu due to the recent gameworks program?

[Topweasel]

Because AMD could just change Mantle, and then Nvidia's implementation would have to be redesigned. That's why they don't give it a shot. It's not worth the risk. AMD specifically said it is controlling Mantle so it can better optimize its own hardware. What part of this makes it hard to understand why Nvidia doesn't want to touch it?

Not to be mean but if you struggle with understanding what an open standard is, how can you call someone on not understanding the highly technical side of the underpinnings of Mantle and how hardware or drivers would be developed to support it?

One seems a lot more complicated and not being EA or AMD one would be hard pressed to be able to pinpoint what implementation would really mean for Nvidia. The other one is a simple phrase with a given value within the context of the thread that is understood by most people.

For all you know Mantle's implementation for Nvidia and their processing units would just be a couple of driver hooks. Remember that Nvidia's unit are still larger and more complicated than AMD's (which is why AMD units end up being better miner GPU's) so it's unlikely GCN could easily do something that Kempler can't. There are a lot of decent reasons for Nvidia not implementing Mantle. But hardware development paths might not be one of them. The simplest and probably more correct statement would be to develop a Mantle driver would mean lending credence to the API, which in turn would push adoption, which in turn would give AMD more clout in the evolution of PC Gaming graphics. They wouldn't want that.

 

[Mand]

If you have to resort to personal insults attacking my credibility, that means you've lost the argument.

This is very simple. You do not slave your hardware to your direct competitor's API, not when their public, clearly expressed goal is to use it to optimize their own hardware. It's not worth the risk, and the payoff is minimal.

Mantle is controlled by AMD. That means it isn't open. My "understanding" is perfectly fine. There's a difference between an open standard and something that's proprietary but whose use is freely licensed. AMD still owns Mantle. They still control what's in Mantle. If they wanted to, they could let Nvidia spend a year working on getting it working on GTX cards, and then change something to sabotage their efforts and Nvidia would have precisely zero recourse. They cannot take that risk.

That's what it means to not be open. One party with complete, unilateral control is the very definition of what it means to be proprietary. Muddy the water about what it truly means to be open all you want, but what it means to be proprietary is very clear. AMD has sole discretion over Mantle, and they said, publicly, that they are going to keep it that way for the express purpose of making Mantle run better on their own hardware.

 

[Topweasel]

If you have to resort to personal insults attacking my credibility, that means you've lost the argument.

This is very simple. You do not slave your hardware to your direct competitor's API, not when their public, clearly expressed goal is to use it to optimize their own hardware. It's not worth the risk, and the payoff is minimal.

Mantle is controlled by AMD. That means it isn't open. My "understanding" is perfectly fine. There's a difference between an open standard and something that's proprietary but whose use is freely licensed. AMD still owns Mantle. They still control what's in Mantle. If they wanted to, they could let Nvidia spend a year working on getting it working on GTX cards, and then change something to sabotage their efforts and Nvidia would have precisely zero recourse. They cannot take that risk.

That's what it means to not be open. One party with complete, unilateral control is the very definition of what it means to be proprietary. Muddy the water about what it truly means to be open all you want, but what it means to be proprietary is very clear. AMD has sole discretion over Mantle, and they said, publicly, that they are going to keep it that way for the express purpose of making Mantle run better on their own hardware.

Again I am not trying to be mean. You simply refuse to understand that your Open doesn't mean the same thing when dealing to patented computer technology. All tech is patented. DDR has dozens of patented tech owned by singular companies and then shared for the standard. DDR is Open in the sense that anyone can use Micron's patents freely. It's Open because Micron and a couple other companies came together an pooled their patents and said that if your making memory to their specs than you can make DDR. It doesn't matter if its one company or a pool of them (and in that sense at least Mantle 1.0 was created by 2 companies (EA)). An Open standard is just a clear outlined patented and owned service or process that has been made open for use to the public. Does not mean open to change.

PDF is an open standard yet a wholly owned extension of Adobe. One that MS has been fighting for the same reasons Nvidia won't support Mantle. Ever try to edit a PDF with anything other than Acrobat? Damn near impossible.

I am not trying to attack you. I am just wondering if you can't understand the simple difference that an Open standard is open for use and not Open for editing? And that not being open for editing doesn't mean it's not open? Why would you expect others to understand the political warfare that goes on between different hardware suppliers. Why they play nice sometimes and dig their feet in others. Specially without having any technical knowledge of the actual service in question. On one side the point is pretty simple on the other without Mantle released its all speculation.

After its released who knows. Nvidia might look at it and decide that it offers their cards so much performance and is so simple to implement in drivers with no change to hardware, that they decide its silly to not support it. Then won't you look silly talking about how people were being too dense to understand why Nvidia wouldn't support it. All you have is PR quotes that you try to gleam technical information from. What Nvidia would or wouldn't have to do and what their reasons for doing or not doing those would have to be decided at that point.

As a side comment. AMD doesn't generally believe in licensing other peoples techs when they have other solutions. But they aren't against using standards established by their competitors when it helps. SSE and it's iterations is a great example of that.

 

[Mand]

You say you're not trying to be mean and that you're not trying to attack me, then proceed to insult me repeatedly.

We're done here.

I'll just quote myself the important bit that you seem to have ignored the first time around:

There's a difference between an open standard and something that's proprietary but whose use is freely licensed.

 

[Hitman928]

insult me repeatedly.

 

[Hitman928]

Also, since no one seems to have actually looked at the link, here is what the people at the open initiative consider and open standard to be. You can see what is included, but also what is not included:

No Intentional Secrets: The standard MUST NOT withhold any detail necessary for interoperable implementation. As flaws are inevitable, the standard MUST define a process for fixing flaws identified during implementation and interoperability testing and to incorporate said changes into a revised version or superseding version of the standard to be released under terms that do not violate the OSR.

Availability: The standard MUST be freely and publicly available (e.g., from a stable web site) under royalty-free terms at reasonable and non-discriminatory cost.

Patents: All patents essential to implementation of the standard MUST:
be licensed under royalty-free terms for unrestricted use, or
be covered by a promise of non-assertion when practiced by open source software

No Agreements: There MUST NOT be any requirement for execution of a license agreement, NDA, grant, click-through, or any other form of paperwork to deploy conforming implementations of the standard.

No OSR-Incompatible Dependencies: Implementation of the standard MUST NOT require any other technology that fails to meet the criteria of this Requirement.

http://opensource.org/osr

 

[Topweasel]

I'll just quote myself the important bit that you seem to have ignored the first time around:

There's a difference between an open standard and something that's proprietary but whose use is freely licensed.

What I and other have been trying to explain is that there isn't. I can't think of what in engineering would be compared to "Open Source". But its not an Open Standard. That's the thing for something to become a standard, it can't be "open" for changes. Someone some where has to close the door at some point. But in engineering even the ones that you think are "open" like lets say HDMI. Is actually highly patented and owned by the companies that partook in its development.

Open in this discussion equals free to use. Open means royalties free for using patented technology. Nothing more nothing less.

 

[3DVagabond]

This has been proven to be factually inaccurate.

3DVagabond, you asked what had been debunked, this claim right here.

You need to try a little harder. You seem to be avoiding A) giving a source, because B) you don't want to say exactly what has been debunked.

For example, are you saying that whole statement has been debunked, or just a specific part? You're trying to make it sound like the entire statement is false, which it's not.

Don't expect others to do your work for you. At the very least you need to be specific in what you are claiming. One liners aren't enough. Imagine if we all were doing that. It would end up like Monty Python's "Argument" skit.

 

[Topweasel]

Also, since no one seems to have actually looked at the link, here is what the people at the open initiative consider and open standard to be. You can see what is included, but also what is not included:



http://opensource.org/osr

Well back to your early post on that. I think all the definitions even the ones on http://en.wikipedia.org/wiki/Open_standard page as outlined by different groups all seem the same on one thing.

The open is referring to the Usage (free) and that the main contention is the use of Standard. Standard is where Mann could make his fight that it isn't a true Open Standard since AMD didn't use a group of impacted parties to help develop the technology. But "Open standard" vs. "Open" in this aspect makes little difference its a free API that Nvidia can use even if they don't want to. Not a proprietary API just because AMD developed it. Physx is a proprietary API.

I would add that AMD has seemed to focus on open tech. They were pushing OpenCL (open) before Cuda (proprietary), Crossfire (platform agnostic) vs SLI (Nvidia chip-sets only). I feel Mantle existence relies heavily on the lack of a solution and them filling void then concocting a new tech to force Nvidia out. I mean just look at A-Sync (inclusion into the HDMI Open Standard) vs. G-sync. It's really hard to see Mantle at this point being a power move, because honestly if they really wanted to they could leave it closed and made it's code be damn near mandatory for console development.

 

[Hitman928]

I mean just look at A-Sync (inclusion into the HDMI [I think you mean DP ] Open Standard)

..

 

[Topweasel]

..

Oops. Yeah Display Port. Not HDMI, which in retrospect is disappointing. As the push to make a PC more user and living room friendly its kind of disappointing that we really won't ever see at least in the short term a form of adaptive refresh rate on TV's. Though its probably better not let TV's establish random refresh rates in terms of Movie playback. One less possible bug to come up and ruin the movie watching experience.

 

[Bateluer]

Proprietary tech and shady 'optimization' processes that bar competitors should make one avoid that company's products like the plague. In this case, Nvidia.

Why would you support that behavior by buying their models?

 

[96Firebird]

Why would you support that behavior by buying their models?

So you can buy AMD's products cheaper.

You're welcome. :biggrin:

 

[3DVagabond]

So you can buy AMD's products cheaper.

You're welcome. :biggrin:

You don't honestly think that nVidia holds down pricing.

 

[Bateluer]

So you can buy AMD's products cheaper.

You're welcome. :biggrin:

AMD is the only reason why Nvidia's mid range cards haven't ballooned to over 600 dollars in price. Nvidia is one company that simply must have a moderately strong competitor to keep them in check.

 

[vanguard27]

From my perspective, after AMD is done with the mantle beta testing and releases the final SDK, it'll be still able to get adapted by Intel or Nvidia. So the option is there for everyone regardless what kind of source it will be. But as for the gameworks, it's entirely a closed source program. Nobody knows better than Nvidia itself about what's in the code. Even when they were asked by a Forbes reviewer about the fact that, if AMD comes to a developer already contracted with Nvidia and offers or suggests anything about the game development or code optimizations for their GPU, will it be considered or not. And the reply from Nvidia was positive and they said that the developer is free to do anything. Now, we don't know anything about the entire gameworks coding. So what if there's something in it that directly disables AMD cards performance? Or Nvidia passively forces the developer to listen and forget what other vendors say just so that they can avoid all the heat and appear apparently 'open to everyone?' Whole thing is a blackbox indeed. They should compete at an open level. No offense to anyone. Just my opinion.

 

[AnandThenMan]

GameWorks is incredibly stupid if you care at all about gaming. It means that Nvidia is the one "optimizing"code for AMD graphics hardware. Does it get any more suspect?

 

[tweakboy]

Whatever is your favorite color. green or red. BTW CCC is horrible.

 

[boxleitnerb]

AMD is the only reason why Nvidia's mid range cards haven't ballooned to over 600 dollars in price. Nvidia is one company that simply must have a moderately strong competitor to keep them in check.

Not entirely true. Higher prices mean less sales which in turn could mean equal or even less profit. Why do you think it is that for every Titan that is sold Nvidia sells hundreds if not thousands of GT720, 730, 740...?
You can only increase prices up to a certain point before it comes back to bite you. Then you simply cannot compensate for the loss of sales with the higher margins.

 

[SirPauly]

So what if there's something in it that directly disables AMD cards performance?

And there’s nothing built in to GameWorks that disables AMD performance? “No, never.”

AMD’s been saying for awhile that without access to the source code it’s impossible to optimize. That’s crazy.”

http://www.forbes.com/sites/jasonev...ut-gameworks-amd-optimization-and-watch-dogs/

I enjoyed the wording crazy! All the whining and yet Watch Dogs ran fine on AMD hardware and if anything this title showed a tangible reason to upgrade for sku's with more ram! Instead of focusing on the strengths of AMD -- the marketing desires to attack a nVidia strength on a strong selling PC title!

Crazy!

 

[NIGELG]

http://www.forbes.com/sites/jasonev...ut-gameworks-amd-optimization-and-watch-dogs/

I enjoyed the wording crazy! All the whining and yet Watch Dogs ran fine on AMD hardware and if anything this title showed a tangible reason to upgrade for sku's with more ram! Instead of focusing on the strengths of AMD -- the marketing desires to attack a nVidia strength on a strong selling PC title!

Crazy!

Your constant use of exclamation points makes your posts irritating and difficult to read through.

Your defense of all things Nvidia is fun to watch though.

''...the marketing desires to attack a nVidia strength on a strong selling PC title!'''.You always say 'the market decides'. So what is the problem?

Gameworks should not factor in a purchase decision because it does not bring any thing much to the table.It showcases Nvidia and their ruthless desire to get ahead at all costs.Gameworks,like PhysX and Stereo 3d, is weak and is trash really.

 

[96Firebird]

Your constant use of exclamation points makes your posts irritating and difficult to read through.

Your defense of all things Nvidia is fun to watch though.

Your constant judgement of other people's posting style makes your posts irritating and annoying to read.

Your defense of all things AMD is fun to watch though.

 

[stahlhart]

Get this discussion out of personal territory ASAP.
-- stahlhart