I guess we know why Nvidia have been milking their current video card architecture for so long. They have been raking in all the money they could from their G80/G92 sales to buy out Ageia.
Printable View
I guess we know why Nvidia have been milking their current video card architecture for so long. They have been raking in all the money they could from their G80/G92 sales to buy out Ageia.
I dont think we will see physics on GPU. Buying Aegia is basicly like saying physics on GPUs will always suck, we need PPU. Maybe one day tho they will merge the 2 dies. But would still be seperate I bet.
AMD and nVidia have basicly been wasting 2-3 years on physics on GPUs without really going anywhere.
lol, i been boycotting them since 2001. My last nvidia card was Ti4200.
Everyone here cant seem to realise that all games released up to date had only used small amount of effect physics because the physx market is very small and game developers dont want to lock them self into a small portion of the games market. You have not seen what a PPU can do, the CPU can handle small physics tasks but for full on gameplay physics the CPU is not enough. This was same when first GPU came on everyone said that its a waste and they should leave it with the CPU but then a killer app came along (Quake) and showed how much more can a GPU deliver.
Because Nvidia bought Agiea we wont be seeing this killer app for quite some time, nvidia is only going to push for general physics effects like they have been doing with DX10.
This could be good and bad. Nvidia's future cards will probably have good physics processing abilities. But the future Nvidia sponsored games will probably have physics options that's exclusively available to Nvidia cards only.
:toast:
YAY!!!! now for some real physics cards to hit the market
That's the thing, with Nvidia's developer relations growing strong, games that will support Nvidia's PhysicX are gonna be a lots, including triple A titles, which will add incentive to buying Nvidia graphics cards --> flawless plan, let's see how good they carry it out.
Finally my time to laugh at those who laugh at me last time, ..."that would never happen - there is no space for an Ageia Physic CIP on PCB" they say. Just look, it happened, I doubt the space will be a problem - they'll just make a bigger PCB (see ATi 3870X2). :D
I notice everyone keeps mentioning. What will Nvidia do with "Ageia." I think its more like. What will Nvidia do with the newly aquired physics processor. Has everyone forgot about 3dfx already. let me put it to you this way with a question.
Are we running Voodoo 8800GT's in our machines?
I dought the name will stay, and if so, Ill be kinda annoyed, caue Ageia is a pretty stupid sounding name, in my opinion.
IF anyone thought space was a problem then thats completely retarded. I mean, look at the freaking 9800GX2 for the love of god. lol
i've just read the article on xbit labs (http://xbitlabs.com/articles/editori...view-2008.html) and it sounds like ageia will stay in business on its own - at least for the time being.
Anandtech article:
http://www.anandtech.com/video/showdoc.aspx?i=3224
Interesting read, with speculation about how NV is going to use Ageia's IP in their future products.
I wonder what AMD is gona do about this....Intel has Havok and Nvidia has Ageia :shrug:
Considering NVIDIA needed a PCB as long as 3870X2's to make a single chip card (GTX and Ultra), and will need two PCBs in the 9800GTX for two chips, I doubt that they can add another chip maintaining a single PCB design.
And that's assuming that the AGEIA chip doesn't need its own RAM...
I have a question...Does anyone actually have a Physics chip in their system? or is all this "it won't make a difference" just speculation?
I am curious b/c I have toyed with getting one for my future build (Going NON-sli)....
In related news to this: NVIDIA completes Ageia buyout.
That was pretty quick, as they completed the buyout 9 days after officially announcing it. I guess they must have been busy with it for quite some time or something.
Nividia + Ageia = meh.
If they begin to utilize the physics tech, great. If they don't, great. I may just be weird, but I don't really see too much potential. As long as graphics continue to get better, I could care less if things do stuff realistically or not.
at least the topic about new ppu hardware is buried. the only usefull thing nvidia acquired is the software layer/sdk tools; the rest is open for speculation ;)
I have a PPU card. And, I played GRAW2, using the setup in sig @ 1920x1200, settings maxed out.
It's gonna take some serious GPUs to do both physics acceleration and rendering at the same time. I mean, some of the physics acceleration in GRAW2 drags frames down drastically (even @ 1600x1200). As the physics in GRAW2 is mainly “effects physics” (and not “gameplay physics”), this drag must be GPU and not CPU related.
For the sake of argument: Try imagining any nVidia GPU running Crysis in DX10, HD, high graphics settings and calculating the physics acceleration of the game - at the same time. I think you get the picture by picture by picture by picture... frame rates :D
Contrary to what Anandtech writes and what has been indicated in various statements on the acquisition (including the press releases from nVidia), I still think we will see a PhysX technology implementation that goes beyond merely using the PhysX software on nVidia GPUs (combined with minor GPU architectural optimizations/improvements). Time will show…
On a side note: To judge from Ageia's US patent application, this acquisition is what they were hoping for, from the very beginning of their effort to bring real-time physics acceleration to 3D environments:
Paragraph [0061].Quote:
[...] It is readily foreseeable that a PPU and a GPU will be combined on a single interface card. That is, both chips will be physically mounted on the same card (AGP or PCI-Express), but not directly interfaced with one another. Ultimately, a single interface card having a directly interfaced PPU-GPU combination is expected, but such a combination is probably a generation away. So too is a combination within a single chip of PPU and CPU functionalities.
That the team at Ageia (in all probability) didn't make a lot of dough through selling now - due to financial difficulties - lies IMO entirely on their inability to understand the diversity of the enthusiast market segment. That is, not only focusing on the gamers, but also the benchers, overclockers, modders and crunchers/folders. Talk about Xtreme disappointment! When Ageia tried to right this huge mistake, it was way too little, way too late.