Print 38 comment(s) - last by 3kliksphilip.. on Feb 20 at 11:59 AM

The physics-intensive Cell Factor: Revolution demo will soon be "No PhysX Card Required"
NVIDIA's purchase of AGEIA leads to a PhysX-on-CUDA port

With the announcement earlier this month of NVIDIA's acquisition of AGEIA, rumours began to fly immediately surrounding the future of dedicated physics hardware -- and it now appears that the PhysX name will live on as a checkbox beside the capabilities of some current and most future NVIDIA GPUs.

During NVIDIA's fourth-quarter financial results conference call, CEO Jen-Hsun Huang responded to several questions about the plans for technology obtained in the AGEIA purchase, revealing that the plan is to port the AGEIA PhysX engine to NVIDIA's CUDA (Compute Unified Device Architecture) C-like programming language.

"We're working toward the physics-engine-to-CUDA port as we speak. And we intend to throw a lot of resources at it." said Huang. "[PhysX on CUDA] is just going to be a software download. Every single GPU that is CUDA-enabled will be able to run the physics engine when it comes."

NVIDIA's choice to run a physics engine on a GPU runs in stark contrast to AMD's assertion in late 2007 that "GPU based physics is dead until DirectX 11." Every NVIDIA 8-series GPU is currently capable of running CUDA applications, and future GPUs will no doubt retain this feature.

The idea of using SLI for more than graphics has been brought up by NVIDIA in the past, so it was no surprise to hear Huang endorsing its further use again. "It might - and probably will - encourage people to buy a second GPU for their SLI slot. And for the highest-end gamer, it will encourage them to buy three GPUs." No mention was made of the use of the upcoming "Hybrid SLI" technology showcased at CES 2008, but an onboard GPU supporting CUDA could theoretically be used as a physics processor while discrete GPUs handle the rendering.

No timeframe for the release of the PhysX-on-CUDA software was specified, but with the PhysX engine to be available to a larger audience, it will no doubt encourage the development of more accelerated physics engines in upcoming titles.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

By omnicronx on 2/15/2008 1:27:27 PM , Rating: 1
Now if it is software based, what makes current GF8 cards and above so special? Why would this not work with say a 7800/7900? lack of processing power?

And really why waste all this time implementing physics into your code, and not take the time to implement a physics chip into the videocard itself. I just don't see software physics being the wave of the feature, its going to have many downfalls.

RE: hmmm
By Vinnybcfc on 2/15/2008 1:48:09 PM , Rating: 2
CUDA was only developed for the Geforce 8, it would be impossible to implement it on lower versions.

It isnt just a case of software, your hardware needs to be compatible with CUDA as well

RE: hmmm
By FITCamaro on 2/15/2008 4:19:39 PM , Rating: 2
The Geforce8 series was the start of the unified shader architecture cards. Previous cards had dedicated pixel and vertex shaders. While pixel shaders can be programmable, vertex shaders are locked to doing a single task. Unified shaders can do the work of either type. This is why it won't run on anything other than the Geforce8 series. It's not possible. The architecture doesn't allow it.

RE: hmmm
By FITCamaro on 2/15/2008 4:24:08 PM , Rating: 2
And I agree that it should go to dedicated physics chips on the GPU itself. But the problem is to do that, you really need a more defined API for it. Otherwise developers will have to code the physics differently for each brand. While yes they already do use some optimizations for each brand of card, they'd have to do even more.

This is where a company like Microsoft comes in handy since they can push the standard out there.

RE: hmmm
By PrinceGaz on 2/17/2008 12:24:29 AM , Rating: 2
The last thing we need at eh moment are dedicated physics-chips. GPUs and CPUs are converging, give them time to do so and PPUs will be irrelevant.

At best PPUs introduce a proprietary new physics-language that will be utilised by at best 1% of game players (I doubt even 0.1% of current PC game players have bought a PhysX card), and therefore a small minority of games.

Over the next year or two, the PC gaming market will advance again as the console world settles into its usual cycles. There will never again be a dedicated PPU, or onboard physics-unit on a graphics-card because it is too small a market to develop for.

But in two or three years, graphics-card shaders will be so flexible that the line between them and general-purpose CPUs will be a little blurred. Your 8 or 16-core CPU will be able to handle a lot of physics, as will your parallel-processor card.

"When an individual makes a copy of a song for himself, I suppose we can say he stole a song." -- Sony BMG attorney Jennifer Pariser

Related Articles
Update: NVIDIA to Acquire AGEIA
February 4, 2008, 5:31 PM
Havok Causes Havoc in GPU Physics
November 21, 2007, 4:46 PM

Copyright 2015 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki