|  | | AMD: "Nvidia PhysX Will Be Irrelevant" | 
Advanced Micro Devices, the world’s second largest developer of x86 central processing units (CPUs) and a leading designer of graphics processing units, said that it hardly regrets about Nvidia Corp.’s decision to disable support of hardware physics effects processing using PhysX API and GeForce GPU or Ageia PhysX PPU in systems where ATI Radeon graphics card is used for graphics rendering. In fact, AMD believes that with the raise of popularity of DirectCompute and OpenCL APIs, proprietary PhysX will soon vanish into oblivion.
“Physics can be a good thing for gamers, but it should be for all gamers. When it’s available for everyone, game developers will be able to make physics an integral part of gameplay, rather than just extra eye candy. This requires a physics solution built on industry standards. That’s why DirectX 11 is such a great inflection point for our industry – DirectCompute allows game physics that can be enjoyed by everyone. There are several initiatives (some open-source) that will deliver awesome GPU-based physics for everyone, using either DirectCompute or OpenCL. Industry standards will make any proprietary standard irrelevant,” said Neal Robison, director of global independent software vendors relationship for AMD, in an interview with Icrontic web-site.
Read the entire article in Xbit-Labs. | | 16 Comments | | | | | ATI is talking out their arse. Due to ATI's lack of working with the developers, PhysX is more likely to gain momentum through the use of the TWIMTBP program. | | | | Further more, the fact that ATI is talking out its arse shows in that PhysX was offered to AMD so the fact that ATI gamers lack physx is ati's fault and not nvidia's. | | | | yeah yeah squall. You forgot that if AMD did join PhysX camp they would have to accept CUDA as well. And CUDA is build AROUND GF SCALAR Stream Processors...not to be effectively used with AMD 5way VLIW superscalar  So sorry, either STREAM enabled PhysX or no physx at all for me. And again...there is a yet unreleased HAVOK OpenCL accelerated...and BULLET too. So no, I don't think that NV PhysX will gain much attention after AMD OpenCL driver and certificate comes out. | | | | ok.....
is well known the fact that ati does not support very well game developers...
if nvidia pays game developers to improve games on pc (nvidia only) may not be a "decent move" BUT at least nvidia is doing something about pc gaming!
im tired of console fanboys saying the consoles are as good as pc's!
NVidia is fighting to change that!
ati does not do SH*T .. just complains!
now the only really PR bad move nvidia made here.. was disable PhySX as an open source...
ati fans are saying now that nvidia is making people to buy their hardware if they want physx!!
bad move....
but then again i wanna see at\amd spending money .. i wanna see them improve .. lool |
Last edited by ObliVioN_; October 4th, 2009 at 01:22 PM..
| Quote | | | | | We aren't going to see any gameplay physics for some time I fear regardless of opencl or direct compute.......its going to take quite a time for DX11 hardware and OS support to become realy established....during that period game developers are not going to exclude all those still using DX9 or 10 stuff from their customer base and are so not going to impliment gameplay physics which would not run on those older systems......we are stuck with eyecandy for the next couple of years. | | | | Quote:
Originally Posted by Unregistered  yeah yeah squall. You forgot that if AMD did join PhysX camp they would have to accept CUDA as well. And CUDA is build AROUND GF SCALAR Stream Processors...not to be effectively used with AMD 5way VLIW superscalar So sorry, either STREAM enabled PhysX or no physx at all for me. And again...there is a yet unreleased HAVOK OpenCL accelerated...and BULLET too. So no, I don't think that NV PhysX will gain much attention after AMD OpenCL driver and certificate comes out. | oh look, another anonymous fool who hasn't got a clue about PhysX or CUDA. | | | | Quote: |
oh look, another anonymous fool who hasn't got a clue about PhysX or CUDA
| No, not really. I just don't remember my pass..so when using different PC I post as Unregistered. I'm Promilus.
And no...I'm not a fool. PhysX uses CUDA. No CUDA capable GPU, no GPU PhysX. And CUDA IS build around G80/G92/GT200/GT300 stream processors structure. Scalar, asynchronous...much unlike 5way superscalar synchronous VLIW of R600/RV770/Cypress. | | | | same old story. nvidia fanboys defending anything nvidia does and vice versa with amd/ati. I'm all for an open standard so that everyone can take advantage of it. locking hardware physics processing only alienates the very people hardware companies want as customers. I have both Nvidia & ATI hardware. The more NVidia lock stuff like this down only sways me further away from them. If ATI did the same thing I'd be moving away from them as well.
Regards, os2 | | | | CUDA is an advanced execution engine which allows several forms of C and with the G300 several other languages as well to execute in shaders.
OpenCL DOES NOT EVEN COMPARE. | | | | I doubt it, seems like ATI is just jelous like usual lol. IMO. | | | | Quote:
Originally Posted by squall_leonhart CUDA is an advanced execution engine which allows several forms of C and with the G300 several other languages as well to execute in shaders.
OpenCL DOES NOT EVEN COMPARE. | Without meaning this as an attack on your position, as a relatively non tech guy I'm wondering if you could explain why open CL doesn't come close. I was sort of hoping it would be a way forward? | | | | Because its maintained by Khronos group, It will be slower to adopt feature sets, there will be infighting, bitching and things cut for no good reason. | | | | So the physics applications won't run any less well on open cl than on cuda it is just the committee approach will slow things down?
But surely that is a small price to pay for a unified approach that everyone can use regardless of hardware?
I really just don't see how a proprietary approach can ever succeed to the customers benefit. | | | | Quote: |
I really just don't see how a proprietary approach can ever succeed to the customers benefit.
| Well for starters, its only propietary until other firms are using it, just like with intel and its first SSE capable cpu's.
imagine how things would be currently if AMD had snuffed intel and gone with their own non-compatible instruction sets.
The Open Format groups just aren't being aggressive enough to keep up with closed source api's, and this unfortunately is a major reason why OpenGL, OpenAL, and OpenCL are bound to never really pick up pace with the majority of Windows only developers.
They will go with the format that is easiest to program for, evolves the fastest, and is constantly updated. Quote: |
Originally Posted by tmurray No--they're separate APIs. There are things you can do in CUDA that aren't possible in OpenCL, and we don't expect that to change going forward (it's not like we're going to stop adding features because OpenCL won't support it). | | | | | Quote: |
The Open Format groups just aren't being aggressive enough to keep up with closed source api's, and this unfortunately is a major reason why OpenGL, OpenAL, and OpenCL are bound to never really pick up pace with the majority of Windows only developers.
| Keep in mind that there is something more than just PC market. OpenCL in theory is able to run the same program on x86, on PPC, Cell, GPU with any mixes you want. Is CUDA capable of the same? No? Well, how can it then be competitive? CUDA - one hardware, OpenCL - plenty of hardware. Yeah yeah...I'm just only stupid master engineer of electronics with speciality of energoelectronics. What can I know...lol | |
|
