-
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 Announces Open Physics Initiative
- Thread starter Final8y
- Start date
Soldato
- Joined
- 28 May 2007
- Posts
- 10,061
I personally think the nvidia gpu in the ps3 is a little disappointing compared with the xbox 360 ati gpu. I don't even think the ps3 can do aa where as the xbox 360 can do 4x. I think the other thing to look at is gpu size and cost where in console's this is very important and atm ati have the upper hand here in a big way.
Sony as a company has always try to segregate itself from the "others". They want to be able to command a premium for their products. Not as far as Apple, but they are always more expensive than other named brands.
Also the amount they have invested in CELL leads me to believe they won't be dropping it for their next console, upping the frequancies and maybe some minor architectural tweaks.
If ATi, which looks about 50% if not more likely, is going to be supplying microsoft with a GPU, then Sony won't go for the same brand GPU. They need to distinguish their product from microsoft's.
Also Playstation still carries a name among console gamers, so what ever of a flop PS3 has been, they still have brand power. Xbox is getting their thou rapidly.
I would like to see nVidia produce a further refined version of their Fermi architecture to Sony, would be epic to see the difference between two somewhat different approachs to the same problem, in-game physics.
Also the amount they have invested in CELL leads me to believe they won't be dropping it for their next console, upping the frequancies and maybe some minor architectural tweaks.
If ATi, which looks about 50% if not more likely, is going to be supplying microsoft with a GPU, then Sony won't go for the same brand GPU. They need to distinguish their product from microsoft's.
Also Playstation still carries a name among console gamers, so what ever of a flop PS3 has been, they still have brand power. Xbox is getting their thou rapidly.
I would like to see nVidia produce a further refined version of their Fermi architecture to Sony, would be epic to see the difference between two somewhat different approachs to the same problem, in-game physics.
Soldato
nice news 
Its not impossible the next playstation will have hardware physx and nvidia GPU again... sony have flirted with it in the past and its something that would be distinct from the rest... I can imagine nVidia pushing for it too to keep flogging the physx horse.
Soldato
- Joined
- 28 May 2007
- Posts
- 10,061
Looks like nvidia won't be in the next round of consoles if this article is true. It says that intel has the contract for the ps4 gpu.
http://www.atomicmpc.com.au/News/136453,intel-beats-nvidia-to-the-ps4-gpu.aspx
http://www.atomicmpc.com.au/News/136453,intel-beats-nvidia-to-the-ps4-gpu.aspx
I think that article is pure specualtion, Intel has nothing right now, absolutely nothing in the market place. Who knows what Larrabee will become, it could fail miserably at graphics and simply become a GPGPU appliance.
Someone at a Sony booth told them all of this, i find that unlikely.
Someone at a Sony booth told them all of this, i find that unlikely.
I'd be suprised... its not what I've heard "down the grapevine" either...
I'm not sure what intel are upto with larrabee and I dunno if they know either - it seems to be a technology without an application - its not really well suited to GPGPU type processing despite what it looks like on paper, its not much cop as a 3D accelerator either and intel doesn't have the mentality to really enter the gaming market - they are far too fire and forget technology wise with an ethos of putting out a polished product that can stand on its own feet, then move onto the next.
Last edited:
CELL was created by STI which is Sony, Toshiba and IBM.
If nVidia are going to push into supercomputing space (the target for CELL) then this may represent a conflict of interests for IBM within the STI partnership. IBM holds much of the IPR to CELL. Sony use it.
Fermi represents a competitor to CELL in that it is targeted to process large amounts of data in parallel - Medical Imaging, Oil exploration, Financial modelling, Defence projects. If Fermi stays in the graphics area only on the play station then it may be tolerated.
Unless.... STI becomes STIN and nVidia join the partnership against Intel and AMD.. but that requires nVidia to partner..
If nVidia are going to push into supercomputing space (the target for CELL) then this may represent a conflict of interests for IBM within the STI partnership. IBM holds much of the IPR to CELL. Sony use it.
Fermi represents a competitor to CELL in that it is targeted to process large amounts of data in parallel - Medical Imaging, Oil exploration, Financial modelling, Defence projects. If Fermi stays in the graphics area only on the play station then it may be tolerated.
Unless.... STI becomes STIN and nVidia join the partnership against Intel and AMD.. but that requires nVidia to partner..
Last edited:
To be honest which physics engine that the developer chooses to use will be by preference & what its being used for if more than one physics is on OpenCL as all hardware should be able to run it.
Soldato
Well, so does larrabee in that respect.
All the major players are looking towards producing powerful and flexible parallel processing units for vector-based computations - nvidia are just the first to market (or they will be come next year). Cell was very much a "jack of all trades", retaining much of the flexibility and control logic of a CPU, while also having powerful vector performance too. In light of Fermi (and presumably larrabee and AMDs fusion) it seems a little weak, but that isn't surprising given that it's well over 3 years old by now. I think cell can take a lot of credit for generating interest in GPGPU-type applications, but it's had it's time.
I don't think that STI can be too protective of cell at this point. They need to produce a successor which can compete with the CPU-GPU combinations that the other major players are developing. Cell was a great proof of concept, but right now isn't strong enough in any particular way to make a real impact.
In this case i don't think time is enough to make up for Intel's inferior approach to graphics processing.