View Single Post
Old 02-02-12, 12:39 AM   #218
Registered User
Join Date: Jul 2003
Posts: 1,526
Default Re: next gen kepler to support dx 11.1, also take a year to rollout all cards

My opinion on the matter from another forum:


To be honest, it's mostly what Nvidia are trying to pull off that pisses me off, as unified standards are always better for developer adoption, but it doesn't make their hardware special in any way compared to the competition, as we have:

1: Direct compute and OpenCL for creating and running programs, wich will overtake Cuda sooner or later, so it's only a matter of time.
2: Direct 3D for graphics wich is owned and controled by microsoft, so don't even think of going there.
3: We have a CPU physics API's as it is, and the original PhysX using GPU's for physics calculations went nowhere(6 games at most), and more and more CPU power becomes available as time goes on anyhow.
4: We have unified standards for 3D glasses, so proprietary standards are going to die and it's not a matter of if, but when.

So now we have Nvidia not wanting to participating in a straight performance fight, and further influencing how "good" their hardware is by adding custom code to some games( I.E. developers paid off), that pushes physics even harder and at the expense of artificially making competitors cards look worse in those titles, and also "recomending" hardware review sites evaluate these new cards with those titles to show them in the best possible light, while under other games, it's gets it ass kicked royally.

AMD could the same game by "enhancing" the shader effects in games even further to leverage all the shading power within the HD7970's, so that gamers have to make a choice between better shader effects in some games, versus better physics effects using Nvidia kepler cards....Nvidia want to play dirty then AMD can do the same in another aspect....But overall, it's still a pathetic way to make your hardware look special and i though we were beyond those days....Remember the DX9 days with SM 2.0, SM 2.0b and SM 3.0?, and all the arguing it caused for years non stop?, i do and don't want to see this happen with physics this time around...



Originally Posted by Exposed

I enjoyed Alice Madness returns with Physx. Some areas ran like a dog with lots of particles going on (ice shattering, etc..) on my 5870, but ran flawlessly and looked much better when I switched over to a GTX 570.

I care more about 3D than Physx, but anything that adds to immersion is welcomed, including increased physics. As long nothing is done artificially to hamper performance on the competition, and Nvidia sponsored games get a boost because of Physx, then I really don't see a problem. People that don't like this (like Skynet) should write letters to AMD to improve their developer support rather than harbor sour grapes.

And my reply:


Then the next Direct X update should also add fluid based GPU physics to the standard, and make it mandatory for all vendors to support in hardware on future GPU releases, and we're back to an even ballgame between AMD and Nvidia, and developer support will come faster this way in many more titles as cards from both companies support it....Playing it their own way, and making it work only on their own cards makes sure the success is fairly limited in the number of titles that support it, just like what happened with GPU based PhysX, and i'm truly dissapointed that Nvidia haven't learned this simple lesson yet...Unified standards win and proprietary solutions lose, always
shadow001 is offline   Reply With Quote