NVIDIA Acquires AGEIA: Enlightenment and the Death of the PPU
by Derek Wilson on February 12, 2008 11:00 AM EST- Posted in
- GPUs
Final Words
Isn't NVIDIA a Hardware Company?
So, rather than being about the PhysX hardware, this acquisition is all about the software. NVIDIA is relying on CUDA as a vehicle upon which to implement AGEIA's PhysX software. This doesn't require specialized hardware beyond a graphics card, we aren't going to be seeing a PPU stuck next to the next generation GPU, and we aren't even going to be seeing a lot of AGEIA's hardware IP inside of NVIDIA's GPU.
Tony Tamasi stated that, while NVIDIA may use bits and pieces of PhysX hardware technology, this is not the goal or focus of the acquisition. The idea is that, as DirectX and Shader Models evolve and as graphics becomes a problem that requires better handling of dependant code, GPUs will inherently become better at physics calculations. With CUDA at the center while shader hardware continues to become more capable, eventually everyone will have the hardware to handle a game built around complex physical interactions. This puts NVIDIA's GPU Computing agenda in a position to effectively benefit the average gamer.
But What of the PPU?
By moving in the direction they are already moving, NVIDIA will eliminate the need for a dedicated physics processor. Especially if AMD can get on board and adopt PhysX. As Graphics cards evolve naturally, they will become better physics processors. In the meantime, the PPU isn't going to just completely vanish. After all, NVIDIA can't port PhysX to their architecture in one night. Our understanding is that current commitments will be met, but beyond that the future is to treat the GPU as a general purpose massively parallel floating point engine that can be used to process physics.
Unfortunately for AGEIA, Cell Factor didn’t turn out to be the GLQuake of the physics world. On the flip side, struggling to sell an overpriced product that didn't offer users a huge tangible incentive isn't a good business model. This move is a logical one for AGEIA as it keeps PhysX relevant even if the stand alone hardware doesn't have much of a future. It also really benefits NVIDIA because they have the opportunity to compete with Intel on physics. Here's to hoping AMD joins forces with NVIDIA on this one.
Facebook
Twitter
RSS
32 Comments
View All Comments
kilkennycat - Tuesday, February 12, 2008 - link
Larrabee...Marketing speak for Intel's GPU-killer wannabee. Now about 2 years out? ROTFL.....recoiledsnake - Tuesday, February 12, 2008 - link
Isn't DirectX 11 supposed to ship with physics support? AMD/ATI has said accelerated physics is dead till DirectX 11 comes out, and I tend to agree with them.http://news.softpedia.com/news/GPU-Physics-Dead-an...">http://news.softpedia.com/news/GPU-Phys...nd-Burie...
http://www.xbitlabs.com/news/multimedia/display/20...">http://www.xbitlabs.com/news/multimedia/display/20...
I am shocked that the author didn't even mention DirectX 11 in this two page article.
PrinceGaz - Tuesday, February 12, 2008 - link
OpenPL would be better than a MS dependent solution like DirectPhysics. AFAIK though, no work is being done towards developing an OpenPL standard.It's a shame that CUDA is nVidia specific as otherwise it might be viable as a starting-point for OpenPL development.
mlambert890 - Thursday, February 14, 2008 - link
Why? I mean really. Why would some theoretical "OpenPL" be "better"? Unless you have some transformational OSS agenda, it wouldnt be.Come back to practical reality. Im sure its important to niche Linux fans or the small Mac installed base that everything be NOT based on DirectX, so go ahead and spark that up.
For (literally) 90+% of gamers, DirectX works out just fine and DirectPhysics *will* be the best solution.
A theoretical "OpenPL" would be the same as OpenGL. Marginally supported on the PC, loudly and often rudely evangelized by the OSS holy warriors and, ultimately, not all that much different from a proprietary API in practical application when put in context *on Windows*.
Griswold - Thursday, February 14, 2008 - link
If I build the church, will you come and preach it?MrKaz - Tuesday, February 12, 2008 - link
"Who is NVIDIA's competition when it comes to their burgeoning physics business? It's certainly not AMD now that Havok is owned by Intel, and with the removal of AGEIA, we've got one option left: Intel itself. "I think you are wrong. Microsoft is the principal actor of this “useful technology”.
If Microsoft adds physics to the DX11 API (or even some add-on to the DX10) with the processing done at CPU and/or GPU level AMD will not lose nothing. In fact any Intel/Havok or Nvidia/Ageia implementation might not be even supported. So there goes the Intel and Nvidia investment down the drain.
Of course you are right by excluding Microsoft now because they didn’t show anything yet while Ageia and Havok have real products. But I think it is all up to DirectX.
Resuming I think it’s all up to Microsoft for physics to succeed because I don’t see different implementations to succeed or to get support from developers.
In the end AMD might even have the final word if get to use the Intel version or the Nvidia version ;) or wait for Microsoft…
Mr Alpha - Tuesday, February 12, 2008 - link
Two things that weren't discussed that should have been:1.
We are already GPU limited in most games, so where exactly are we supposed to get this processing power for physics from? Will I in the future have to buy a second 8800GTX for $500, instead of a PhysX card for $99, like I can do now?
2.
Hasn't there been some rumours about Direct Physics, where Microsoft does vendor neutral, GPU accelerated physics?
spidey81 - Tuesday, February 12, 2008 - link
I could be mistaken on this, but hasn't Nvidia decided to start putting onboard graphics on ALL motherboards with their chipsets? Wouldn't that be a great way to have a second GPU to handle physics. Albeit, it may be an underpowered GPU, but it may work well enough to offload physics from the CPU. Like it was pointed out, beats having that horsepower sitting there going to waste.7Enigma - Tuesday, February 12, 2008 - link
I had the same question until I re-read the "official" position of Nvidia. They don't plan (for mainstream market) to integrate the chip into upcoming gfx cards, rather its a technology they can keep in-house until the time arrives where it could be beneficial.At some point if/when physics becomes the norm instead of the feature, I could see the company offering a hybrid gfx chip as the higher-end part(s). Just imagine in 5 years the current AMD 3870X2 was a single high-end gpu, with an attached physics chip (using this as an example since IMO it had been the most effective dual chip solution in real-world performance to date, Nvidia's previous try not so much). That would be practical, again as long as the software is coded for it.
Here's hoping THAT is the future.....I don't want ANOTHER piece of equipment that requires upgrading at every new build.
wingless - Tuesday, February 12, 2008 - link
Intel is going down in 2010! Thats all I'll say about that lol.Seriously though, if AMD manages to get on board with this PhysX thing, their Fusion CPU's and CrossfireX will make a helluva lot of sense. Think about GPU physics on the processor itself, motherboard north bridge, and GPU all adding their powers together to run games with tons of physics. An AMD/ATI or AMD/Nvidia box would make more gaming sense than using Intel CPU/GPU's. This could be a big help to the ailing AMD right now...