Caporegime
- Joined
- 8 Jul 2003
- Posts
- 30,080
- Location
- In a house
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD says PhysX will die
- Thread starter LoadsaMoney
- Start date
Associate
- Joined
- 17 May 2004
- Posts
- 1,585
Well considering that Nvidia have made Physx available to anyone that wants to implement it on there hardware, including AMD, not sure how closed and proprietary it is.
This is already available on AMD GPU's via a 3rd party development as AMD have chosen so far not to do anything with it.
Nvidia said back in March that would make the Physx API available to anyone that wanted it for free, including other GPU developers.
As such I would say that Physx is no more closed and proprietary then Havok, which interestingly AMD are working with. It really wouldn't be that hard technically for AMD to get Physx working on there GPU.
Smells of the usual marketing stuff in retaliation to the recent announcement of EA and 2KGames are implementing Physx to me.
Certainly the dedicted PPU chip is dead and buried, Nvidia announced almost straight away that would not develop the seperate chip further but would port the API to CUDA and run on GPU instead.
This is already available on AMD GPU's via a 3rd party development as AMD have chosen so far not to do anything with it.
Nvidia said back in March that would make the Physx API available to anyone that wanted it for free, including other GPU developers.
As such I would say that Physx is no more closed and proprietary then Havok, which interestingly AMD are working with. It really wouldn't be that hard technically for AMD to get Physx working on there GPU.
Smells of the usual marketing stuff in retaliation to the recent announcement of EA and 2KGames are implementing Physx to me.
Certainly the dedicted PPU chip is dead and buried, Nvidia announced almost straight away that would not develop the seperate chip further but would port the API to CUDA and run on GPU instead.
The problem isnt that ATI cards cant run it, but nvidia set all the standards on it. This will lead to PhysX being optimised for there cards only.
Soldato
- Joined
- 24 Oct 2002
- Posts
- 6,242
- Location
- Portsmouth
To be honest can you expect AMD to say anything else? They're not likely to endorse it are they.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Has anyone played the latest physx game, Mirror's Edge, too much time making a few material banners and absolutely no time spent putting in a story, or gameplay. Woo theres bits of paper on this windy rooftop, wonder who keeps throwing away paper on the rooftop, and how it stays up there until I run over it..........
Also, the fact that the "physics engine" is there to make the bits of paper move right and the glass shatter, but the collision engine is woeful, and 90% of the jumps you don't have to get perfect but get within a certain distance and you auto hook on. So no gameplay whatsoever, like every other Physx title, but debrie.
Now I guess it will be a day or two before Pottsey, and the other guy now, forget the name, who come in and say this was never going to be a big PPU game, but the next one will be great. AS they/he have been saying for the past 2 years for every last PPU game to be released. Sooner it dies the better.
Their slogan should read, "Physx cards, for those that like **** spread over the floor in their games", or more accurately "Physx cards, for those that like **** games".
Also, the fact that the "physics engine" is there to make the bits of paper move right and the glass shatter, but the collision engine is woeful, and 90% of the jumps you don't have to get perfect but get within a certain distance and you auto hook on. So no gameplay whatsoever, like every other Physx title, but debrie.
Now I guess it will be a day or two before Pottsey, and the other guy now, forget the name, who come in and say this was never going to be a big PPU game, but the next one will be great. AS they/he have been saying for the past 2 years for every last PPU game to be released. Sooner it dies the better.
Their slogan should read, "Physx cards, for those that like **** spread over the floor in their games", or more accurately "Physx cards, for those that like **** games".
Last edited:
AMD is running scared... if it weren't the case they'd be putting less spin on it - physx is far from as closed and proprietary as they are making out - infact for a very reasonable fee you can license the source for your product and easily implement the API for no money at all even for commercial useage.
While I'd like to see an open standard - fact of the matter is physx is by far the best implementation on the market with any kind of um forward momentum.
While I'd like to see an open standard - fact of the matter is physx is by far the best implementation on the market with any kind of um forward momentum.
drunkenmaster said "AS they/he have been saying for the past 2 years for every last PPU game to be released. Sooner it dies the better."
Why do you often make stuff up? That is a complete lie on your part yet again. I do not say every time a PPU game comes out the next one will be the big one.
I also noticed you did the normal vanishing trick a few physics threads ago. Once again you where both proven wrong and asked for evidence and once again you just vanishing from the thread.
As it stands right now Physx on the GPU is far better then the CPU. Why do you think all the big devlopers are singned up for it? Why do you think all the big developers are signing up for it?
drunkenmaster said "90% of the jumps you don't have to get perfect but get within a certain distance and you auto hook on."
That's done for game play reasons as pixel perfect platform jumping games tend to get very frustrating for most gamers and don't sell well. Not played it so perhaps they have gone too far, but not being pixel perfect is understandable.
drunkenmaster said "Woo theres bits of paper on this windy rooftop, wonder who keeps throwing away paper on the rooftop, and how it stays up there until I run over it.........."
Most normal large city's have rubbish on the rooftops and the rubbish blows from rooftop to rooftop. Your complaining because they made the rooftop more like real life?
Why do you often make stuff up? That is a complete lie on your part yet again. I do not say every time a PPU game comes out the next one will be the big one.
I also noticed you did the normal vanishing trick a few physics threads ago. Once again you where both proven wrong and asked for evidence and once again you just vanishing from the thread.
As it stands right now Physx on the GPU is far better then the CPU. Why do you think all the big devlopers are singned up for it? Why do you think all the big developers are signing up for it?
drunkenmaster said "90% of the jumps you don't have to get perfect but get within a certain distance and you auto hook on."
That's done for game play reasons as pixel perfect platform jumping games tend to get very frustrating for most gamers and don't sell well. Not played it so perhaps they have gone too far, but not being pixel perfect is understandable.
drunkenmaster said "Woo theres bits of paper on this windy rooftop, wonder who keeps throwing away paper on the rooftop, and how it stays up there until I run over it.........."
Most normal large city's have rubbish on the rooftops and the rubbish blows from rooftop to rooftop. Your complaining because they made the rooftop more like real life?
If AMD was scared they'd just bite the bullet and licence it. In fact they probably have it running on their cards in their R&D department just in case.
The fact is Physx as done knack all and I own a PPU, so its not like I don't know what its about. DX11 and Opencl is just over 12 months away so I'm not sure how much damage NV can do with Physx in that timescale.
Look how long its taken physx to come to anything at all - 12 months plus lead in time is going to mean a LONG time til DX11 and opencl come to anything...
But game developers are finally starting to look at accelerated physics and physx is currently the most promising for most so if it becomes adopted before DX11 and open CL they might not have much impact - if nVidia do what they've done in the past and push developers hard including making dedicated support teams available for implementing these features you can pretty much kiss any other physics implementation good bye.
But game developers are finally starting to look at accelerated physics and physx is currently the most promising for most so if it becomes adopted before DX11 and open CL they might not have much impact - if nVidia do what they've done in the past and push developers hard including making dedicated support teams available for implementing these features you can pretty much kiss any other physics implementation good bye.
Permabanned
- Joined
- 19 Jun 2007
- Posts
- 10,717
- Location
- InURmama
TOTAL BS. 
http://forums.overclockers.co.uk/showthread.php?t=17950621&highlight=PhysX+username_helmutcheese
2 for now but how many will follow.
I'm beginning to think ATI cards cannot do it as Nvidia offered it to them freely.
http://forums.overclockers.co.uk/showthread.php?t=17950621&highlight=PhysX+username_helmutcheese
2 for now but how many will follow.
I'm beginning to think ATI cards cannot do it as Nvidia offered it to them freely.
I'm wondering - I'm far far from an expert and could be completely wrong (my involvement is more software API implementation than the back end) but a quick look at the way physx works on a GPU and the way ATI's streaming processors work (super scalar) suggests than a 4870 would be similiar performance to a 9600GT for physics processing and couldn't sustain the workload without a moderately higher rendering performance hit - compared to nvidia cards.