Topic: NVIDIA to aquire AGEIA

[idolminds]

SANTA CLARA, CA — FEBRUARY 4, 2008—NVIDIA (Nasdaq: NVDA), the world leader in visual computing technologies and the inventor of the GPU, today announced that it has signed a definitive agreement to acquire AGEIA Technologies, Inc., the industry leader in gaming physics technology. AGEIA's PhysX software is widely adopted with more than 140 PhysX-based games shipping or in development on Sony Playstation3, Microsoft XBOX 360, Nintendo Wii and Gaming PCs. AGEIA physics software is pervasive with over 10,000 registered and active users of the PhysX SDK.

"The AGEIA team is world class, and is passionate about the same thing we are—creating the most amazing and captivating game experiences," stated Jen-Hsun Huang, president and CEO of NVIDIA. "By combining the teams that created the world's most pervasive GPU and physics engine brands, we can now bring GeForce®-accelerated PhysX to hundreds of millions of gamers around the world."

"NVIDIA is the perfect fit for us. They have the world's best parallel computing technology and are the thought leaders in GPUs and gaming. We are united by a common culture based on a passion for innovating and driving the consumer experience," said Manju Hegde, co-founder and CEO of AGEIA.

Like graphics, physics processing is made up of millions of parallel computations. The NVIDIA® GeForce® 8800GT GPU, with its 128 processors, can process parallel applications up to two orders of magnitude faster than a dual or quad-core CPU.

"The computer industry is moving towards a heterogeneous computing model, combining a flexible CPU and a massively parallel processor like the GPU to perform computationally intensive applications like real-time computer graphics," continued Mr. Huang. "NVIDIA's CUDA™ technology, which is rapidly becoming the most pervasive parallel programming environment in history, broadens the parallel processing world to hundreds of applications desperate for a giant step in computational performance. Applications such as physics, computer vision, and video/image processing are enabled through CUDA and heterogeneous computing."

AGEIA was founded in 2002 and has offices in Santa Clara, CA; St. Louis, MO; Zurich, Switzerland; and Beijing, China.

The acquisition remains subject to customary closing conditions.

More details about the acquisition will be provided during NVIDIA's quarterly conference call, to be held on Wednesday, February 13, 2008 at 2:00 PM, Pacific Time. The Company's prepared remarks will be followed by a question and answer period, which will be limited to questions from financial analysts and institutional investors. To listen to the conference call, please dial 212-231-2901; no password is required. The conference call will also be webcast live (listen-only mode) at the following Web sites: www.nvidia.com and www.streetevents.com.

Replay of the conference call will be available via telephone by calling 800-633-8284 (or 402-977-9140), passcode 21354792, until February 20, 2008. The webcast will be recorded and available for replay until the company's conference call to discuss its financial results for its first quarter, fiscal 2009.

About AGEIA
AGEIA Technologies, Inc., is the industry leader in gaming physics technology. AGEIA's PhysX software is widely adopted with more than 140 PhysX-based games shipping or in development on Sony Playstation3, Microsoft XBOX 360, Nintendo Wii and Gaming PCs. AGEIA physics software is pervasive with over 10,000 registered and active users of the PhysX SDK. AGEIA is also credited with developing the world's first dedicated hardware physics processor, the AGEIA PhysX processor. The company, headquartered in Santa Clara, Calif., is privately-held. For more information visit http://www.ageia.com.

About NVIDIA
NVIDIA is the world leader in visual computing technologies and the inventor of the GPU, a high-performance processor which generates breathtaking, interactive graphics on workstations, personal computers, game consoles, and mobile devices. NVIDIA serves the entertainment and consumer market with its GeForce® products, the professional design and visualization market with its Quadro® products, and the high-performance computing market with its Tesla™ products. NVIDIA is headquartered in Santa Clara, Calif. and has offices throughout Asia, Europe, and the Americas. For more information, visit www.nvidia.com.

We all sort of figured the only way a physics processor would work is if they bolted it onto a graphics card. No one really wants to go and spend $150 for a PPU.

[Quemaqua]

Yeah, I think we all saw this coming a mile away.

[MysterD]

I think what we figured might happen finally did.

At least it won't come to 500 different boards to process 500 different things -- a board for graphics, one for physics, one for shadows, etc etc. One graphics board is enough, thanks.

[gpw11]

One graphics board is enough, thanks.

Or, you know...three if Nvidia gets their way.

[Jedi]

It's the same with ATI/AMD, their bringing out a 3 card crossfire-whats-it.
On the other hand both are putting dual GPUs on some of their cards so it's like having two cards.

[Xessive]

Hmm, I was actually expecting Intel to pick up AGEIA. I'm kinda glad they got picked up by somebody since this will mean their PhysX technology will be integrated rather having to buy a separate card for it.

[scottws]

This should be a good thing.  I remember reading that the separate physics processing was problematic because then it all had to be sent over to the graphics card for rendering the positions of the objects, slowing things down.  If it's on the same card as or even integrated with the GPU, this shouldn't be a problem.

[Cobra951]

Exactly.  It needs to be anchored at one point (the CPU) or the other (the GPU).  I thought Intel was going to acquire them as well.  The GPU end may work out better, but it would take a different approach to program physics, to put it mildly.  Without the calculations close by on the CPU side, the mainline code is going to need to constantly query the graphics/physics card to stay on the same page with the real-time object actions and interactions (which the main code would not know about directly, since they've been offloaded to the graphics/physics computer).  This is very doable, and hopefully done through a full-featured, well-documented API.  Tell the card to handle something, then get back the results from the card.  All the hard work was offloaded.  The problem I envision is the paradigm shift for game programmers.  They always have to go through those anyway, but it still will take a bit of time after this animal hits the market before it's being used to full advantage.  If the process is similar to the current approach of separate physics and graphics cards (with much faster results after integration) then whoever's already on board will have the jump start.

The more I think about it, the more I like it.  Good tidings.

[Pugnate]

In a few years we are going to see GPUs and CPUs as one and the same.

[Xessive]

In a few years we are going to see GPUs and CPUs as one and the same.

Wasn't that kinda the case before hardware rendering? Back when the CPU just did the whole thing with Duke Nukem 3D Those were simpler times hehe

I wonder what will happen when we finally reach the stage when hardware can generate graphics virtually equivalent to reality. Will some jerk come up with a technology that renders at a higher definition than reality?

[Cobra951]

I can't see the GPU functinality back in the CPU anytime soon, really.  With multicore CPUs, it's definitely possible, but with the current architecture it's immensely faster to have the GPU and its memory on the final hardware stages before the monitor itself.

[Pugnate]

Well they are basically going to be make something that can do both. That seems to at least be the plan within the next two years. It is one of the reasons why AMD bought ATi.

[Cobra951]

They would need to improve throughput to the integrated monitor ports tenfold to begin with.  They would need to have something like dual-ported RAM (for independent unfettered access from both the CPU and the display hardware) or separate dedicated graphics RAM.  In other words, it's much more than the core processing that needs to be reworked, in order to have an integrated solution that trumps the best from Nvidia.

[scottws]

Will some jerk come up with a technology that renders at a higher definition than reality?

Maybe, but it doesn't matter.  The human eye would be the limiting factor.  Basically, it can't ever look better than reality unless some tech/body mod for "HD vision" comes out.

[Cobra951]

Heh.  Yeah, it only needs to be as good as our eyes.  I think we're close right now.  I'm still floored by the technical achievement in Burnout Paradise, and that's just on a console, not a top-tier PC.  Improvements ahead are incremental, not revolutionary--like going from sprites to 3D models.  I do think diminishing returns are reality already.  Once you can do a fully convincing scene in sync with the display hardware, you have arrived.  Beyond that, it's all up to artistry.

[scottws]

I would have asked videocard makers what they would do for business if we ever got to that point, but I guess AMD's aquisition of ATi and nVidia's acqusition of Ageia already answer that.  It will take a ton of computing power to calculate physics.  They aren't even close to mimicking the real world physics yet.

[Quemaqua]

I think real-world physics will be the next step, and beyond that, also liquid dynamics and sound physics.  In other words, once real physics gets to a good point, I think we'll see a lot more realistic stuff written for water and other such substances, and then eventually, once everything else has been done, they'll start moving into computing dynamic sound based on the programmed dimensions of an object.  At least, the little happy part of my head that likes to fantasize about stupid shit thinks that would be neat to see someday, despite the fact that it's probably all but completely impossible.

[scottws]

The problem with real world physics is you are talking about interactions on an atomic level, for different types of atoms.  Computing power is just no where near being able to do that.  I was just reading an article in Wired about the premier CGI special effects guy, and he hasn't even been able to make a algorithm that can mimick a wave properly, meaning make it interact with the wind, sand, boat hulls, etc.  And that's just water, air, and a couple other surfaces.

[gpw11]

In a few years we are going to see GPUs and CPUs as one and the same.

I've heard that a few times, but always thought it applied more to ultra-portable devices, cell phones, and devices that don't really have any major graphical functionality yet (HD-DVD/BRD players?) The benefits wouldn't be for home systems with well defined architecture and a need for cutting edge number crunching power.  The other interpretation of this I've head is that they'll remain separate devices, but take on some of the traits of each other.  So, you'll have a multicore CPU where a few of the cores are specialized and dedicated to act as sort of half-assed GPUs, and you might have a GPU with a very weak and specialized CPU on the same board.  They'd be used in conjunction still, but the specialized cores would take on some of the smaller tasks of the other part so as to cut down needless transfer time between the  two.  I think.  I probably got that wrong.

As for true photo-realism, I agree with Cobra and think the diminishing returns will be the barrier we will never overcome when it comes down to that.  It's hard for a company to justify spending a  shit-ton of cash developing art assets in order to make something look only marginally better.  Procedural synthesis could help with this though.

[scottws]

I disagree with that last paragraph.  Eventually someone will make tools that allow easy manipulation and creation of photorealistic assets, just like how tools are created to help with what's created today that would have been unimaginable to create in the Doom era for instance.

[MysterD]

GeForce 8, 9, and GTX-200 series owners can expect a driver coming up August 12th that will officially bring PhysX features into their NVidia GeForce cards.

NVIDIA PhysX Previews [August 06, 2008, 3:09 pm ET] - Viewing Comments

NVIDIA plans on launching patches on August 12 to enable its GeForce 8-series and later graphics cards to run PhysX physics modeling on the GPU. Presumably an NDA about this expired today, as there are new previews of this new PhysX support on several sites, offering background, hands-on impressions, and benchmarks. You can find these previews on Custom PC, Elite Bastards, HotHardware, and The Tech Report. Here's a summary of the situation from the latter preview:

    On August 12, Nvidia will release new graphics drivers that will allow owners of most GeForce 8, GeForce 9, and GeForce GTX 200-series cards to use PhysX acceleration without spending a dime. Along with the drivers will come a downloadable PhysX software pack containing free Unreal Tournament 3 maps, the full version of NetDevil's Warmonger, a couple of Nvidia demos, and sneak peeks at Object Software's Metal Knight Zero and Nurien Software's Nurien social-networking service. Nvidia provided us with early access to the pack, and we've been testing it over the past couple of days.

[sirean_syan]

... you dug up this thread to post that?

Huh.

[Quemaqua]

I just wonder if there's anyone who actually cares.  It just seems so... unexciting.

[idolminds]

I care. A little. Ill at least be able to play those special UT3 maps.

[scottws]

I wonder how much it will affect performance, running those physics calculations on a GPU that wasn't designed with that in mind.

[MysterD]

... you dug up this thread to post that?

Huh.

B/c it's NVidia and PhysX-related.

This'll be the first NVidia GeForce driver to include the new PhysX-support from NVidia (since they took over Aegia).

I wonder how much it will affect performance, running those physics calculations on a GPU that wasn't designed with that in mind.

Hopefully, not too much.
I am curious, though...

[Cobra951]

I vaguely remember a story about using an Nvidia card to process audio data at 10X the speed of normal programming for the main CPU (including math coprocessing).*  I have a feeling that videocards are so much more powerful for work related to screen output that it's a worthwhile tradeoff.  No external bus or communication with external processors.  Local computations and output directly to the display.

Edit:
* I think this was it.

[MysterD]

This is interesting:
https://blogs.nvidia.com/blog/2018/12/03/physx-high-fidelity-open-source/

NVidia's PhysX has now gone Open Source.

[gpw11]

That's big news....right?

[MysterD]

The most popular & well known Physics Engine (PhysX) going Open Source....of course it is.
Nvidia doesn't normally have Proprietary stuff go open source - so yeah, it's a big deal.

Probably really going to help w/ AI, robotics, and vehicles here - since NVidia already is getting their stuff everywhere in those departments.

Now, also - anyone can use the code and make stuff w/ it.

Also, this could really help older games that actually (for some reason) required PhysX package installations (think GR:AW games, ME2, Sacred 2, and Velvet Assassin come to mind - there are others). Doesn't effect Nvidia users too much (easy enough to install Legacy PhysX drivers, if you are an Nvidia user), but this is especially important for those who did not utilize Nvidia products - i.e. any AMD users. Now, if someone wants, they can take the Open Source from PhysX, mod the PhysX code and files if need me, mod the game, and make 'em work much easier - let the mod out and we're good to go.

Otherwise, like it has been before this open source announcement - one would have to install NVidia's Legacy PhysX software package or the old AEGIA PhysX package; or find the files needed to make this crap work & mod it that way...hopefully. Cross your fingers that it works.

Always nice to see Propriety software and any games go Open Source at some point - especially if they're trying to keep software and/or games alive and preserved, lots of years down the road.