AMD Mantle API [updating]

All those vendors existed then too; I quoted you for a very specific reason. Maybe you should read your own words again: one vendor getting faster? Fine. Another vendor getting faster? Your insipid posts in this thread.

I see a commonality to your post history, and even if you're not aware of your own bias, it still exists.

 

Everyone is biased, including you and including AAA game developers too. Would I be happy to see NVIDIA (or other IHV's) push and promote their own IHV-specific graphics API? No way. That would be sheer lunacy IMHO.

 

Isn't it a stab at Nvidia and the Batman Arkham controversy ?

Cheers

 

The difference is this: I"m not trying to hide my bias behind "oh but that would require effort and be unfair (to my preferred IHV.)"

The expenditure of effort is up to the engine builder. If they find the effort worth their while, then it's a net gain for the users (as you so eloquently put it above in your quote.) If they don't find it worth the effort? Then nothing is gained, and nothing is lost.

 

This is splitting hairs. Pick other examples: physx, cuda in jc2, cuda in rage, etc.

It's also not clear that Mantle is strictly for performance gains (not saying it isn't though).

 

I too believe this to be the case. Performance is assuredly part of it, but that performance can be "traded off" for other things -- and Mantle may expose many more things that can be traded.

 

The BF4/Mantle slides mention unused GPU functionality as a current problem, and that Mantle allows full access to the graphics hardware capabilities.

http://www.frostbite.com/connect/#battlefield-4-frostbite-mantle

Slides 13 and 18.

 

Are you really going with a deal with it reply. You know how much we like those

ps: if nv do their own api I could certainly see them try to use their influence and money to get some games to support that api exclusively.

 

I was primarily recycling that quote of his and feeding it back to him. It's a bit of a low blow, but hey -- if your stated opinion is that other IHV's can have performance benefits and it's "good" for the consumer, then it's intellectually dishonest to deny another IHV the same.

If it were a purely exclusive decision (i.e. your entire game doesn't work on ANY other IHV) then it may be a problem. This was the situation with GLIDE games, way back in the day. At this point in time, I'm aware of nobody who is going pure-Mantle-only.

 

Yes, that's exactly the kind of shitty thing nVidia would do, but not something I fear from AMD. For a long time AMD has had a "do no harm" policy with their dev relations. nVidia's ethical lapses in that area are the primary reason I don't buy their hardware.

 

So the quote from the pcars developer is bogus ?

So seeing as you are sort of locked in to nvidia because of 3dvision what are you going to do ?

 

Most games releasing today choose not to support Deferred Contexts because the gains are just not there. You can usually achieve more performance by making sure your render thread is solely responsible for making DX API calls, and offloading everything else (including buffer processing) onto other threads. On today's APIs this approach is usually the best thing you can do to leverage multi-threading.
To this date only two games that I know off support Deferred Contexts (Civ5 and AC3).

 

Cross my fingers that either Nvidia releases a Mantle equivalent or Microsoft pull their finger out and roll Mantle like efficiencies into a new version of DX that is compatible with Windows 7 (lol). In either case I'd be a winner but if neither happens I've lost nothing the way I see it (as Albuquerque and Brad are arguing above).

Same goes for TrueAudio.

I must admit though that given the tradeoff of Lightboost/3DVision 2 + G-Sync or TrueAudio + Mantle, I'd still go with the Nvidia option. My love of stereoscopic 3D gaming must be well known by now

 

Bad developer decisions hurt consumers. Inefficient APIs hurt consumers.

Consoles use proprietary APIs. Does that hurt consumers? I'd argue that consumers actually hugely benefit from them.

Customers pay for the product as offered. They can ask for whatever they want, but buying A doesn't put developers under any obligation to provide B.

 

Consoles is not a valid comparison here obviously. There is no competition or alternatives.

 

That is true. You can still find parallels though. When you buy a PS4 you're buying into an entire ecosystem not just the hardware. You can look at GPUs the same way.

There's no rule that says two competing products must give you the exact same experience - that's the definition of product differentiation. PS4 and Xbone won't do that for you so why should an AMD and nVidia GPU be any different?

We've been conditioned to think that graphics vendors should only compete on price and performance. In reality that's a very arbitrary limitation. I'm not arguing in favor of market fragmentation, just saying that the expectation of 100% feature parity between competitors is kinda silly.

 

I don't get how Nvidia users are being hurt by Mantle? They are getting the same experience as they had before so the only hurt is the hurt they feel at AMD owners getting a better experience. Isn't that what PhysX is supposed to do also?

Mantle doesn't lock anything out for Nvidia users so it can't be compared to the Batman AA fiasco.

 

I have not read in this thread, I think, why or how mantle will bring benefit to the PC gaming in general.

 

Then read the thread? Or my slides: http://www.frostbite.com/connect/#battlefield-4-frostbite-mantle

 

slide n. 13: pc challenges. How to solves it? By a proprietary solution that benefits one of the pc hardware vendor, and by saying that it would be extremely bad if other IHV's tried to pursue their own IHV-specific graphics API. ( )