NVIDIA To Buy AGEIA

The two companies announced today that NVIDIA will acquire PhysX maker AGEIA; terms were not disclosed. The Daily Tech is one of the few covering the news to go much beyond the press release, mentioning that AMD considered buying AGEIA last November but passed, and that the combination positions NVIDIA to compete with Intel on a second front, beyond the GPU — as Intel purchased AGEIA competitor Havok last September. While NVIDIA talked about supporting the PhysX engine on their GPUs, it's not clear whether AGEIA's hardware-based physics accelerator will play any part in that. AMD declared GPU physics dead last year, but NVIDIA at least presumably begs to differ. The coverage over at PC Perspectives goes into more depth on what the acquisition portends for the future of physics, on the GPU or elsewhere.

Comments

off on a tangent

[User 956]

The Daily Tech is one of the few covering the news to go much beyond the press release, mentioning that AMD considered buying AGEIA last November but passed

Well, that's because they were pondering a similar strategy to Microsoft, and were going to buy Yahoo.

Re:off on a tangent

[ma1wrbu5tr]

Here's a strange tangent... or two

It's almost like some bizarre comic.

Lets imagine that AMD and ATI teamed to to be the Super Friends.

And Intel and nVidia are the Legion of Doom.

Now, let the battle for the universe begin.


At least that's how I feel when I read ./ers comments sometimes.
We geeks tend to take ourselves entirely too seriously,

Grammatical and spelling errors are bonuses.

Re:off on a tangent

[mabhatter654]

except Intel doesn't think they need Nvidia... that's why they've got nearly all the notebook vendors pumping out crappy built-in graphics that just barely run Windows Vista. ATI saw the writing on the wall and got themselves bought by AMD. Now AMD battles on CPUs, integrated graphics, and high end graphics... Intel can never buy Nvidia because they'd be instantly sued. Nvidia overpriced themselves, even with all the work they did for AMD, and the matching logos... stock holders were just too rich for AMD.

This makes Nvidia the "odd man out" because they don't make processors. Both Intel and AMD have integrated solutions and obviously want physics processing on the CPU so that they can sell 7 core 3.21GHz processors. NVidia has to break the mold if they want sales... they got shunned the last round of consoles for IBM and ATI, and Microsoft pretty much let ATI write the book for DX10 this round. NVidia + Ageia only makes sense if they'll make an open source console that runs either AMD or Intel CPUs. Games would need to run flawlessly, without "installing" just like a console. There's a hole for PC gaming right now... Apple's not filling it (they think it's stupid) Wintel is not helping (Microsoft only wants Vista gaming, and Intel wants to sell integrated graphics) so a well done Linux console could help... but there's too much IP in the way to make it happen.

Re:off on a tangent

[greazer]

they got shunned the last round of consoles for IBM and ATI, and Microsoft pretty much let ATI write the book for DX10 this round. Last I checked, the graphics in PS3, a.k.a. RSX, was NVIDIA designed.

Re:off on a tangent

[Calinous]

What about nVidia and Sun? Sun has plenty of intellectual property in processors area - too bad its current processors are optimised for many threads of lower performance

Re:off on a tangent

[Type44Q]

"NVidia has to break the mold if they want sales... they got shunned the last round of consoles" The RSX "Reality Synthesizer" in the PlayStation 3 is a modified nVidia G71.

Re:off on a tangent

[default luser]

But Sun doesn't have an x86 processor, and this is the key.

Nvidia needs a an x86 processor to compete. Sure, Nvidia could just adapt their GPU architecture and expand the language to make a general-purpose VLIW processor. They could package it and sell it as an Itanium competitor. But nobody wants to use a non-x86 chip in mainstream markets, and that's where the long-term money is.

This is why Windows, Linux, Solaris, BSD, and now even OS X run on x86: if your OS has redeeming or unique qualities, more people will buy based on OS features alone if your hardware platform is agnostic.

And herein lies two problems: one, while you can make x86 processors without a license, you are constantly in danger of litigation from Intel's massive patent portfolio. In the last two decades, every x86 chipmaker has eventually negotiated a cross-license agreement with Intel. The other problem is, it is hard to build a new x86 processor from-scratch. Thus, a takeover bid for an x86 processor manufacturer is likely the best way to solve Nvidia's problem; they get a license to keep Intel at-bay, and a solid starting point.

I'm thinking Via, personally. Their sales have slumped in the last year, and they've stopped making Intel chipsets. In fact, Intel has been bullying poor Via for the last year, offering a new Intel chipset license if they just stop manufacturing CPUs. Either Nvidia will buy Via, or Via will spin-off their processor division for some cash. Thanks to the Intel cross-license Via purchased along with IDT, their processor arm is a goldmine in the long-run.

Re:off on a tangent

[mabhatter654]

both Xbox and Wii were designed/manufactured by ATI... you see how PS3 is selling? nVidia got the short straw this round.

Re:off on a tangent

[mabhatter654]

they're rolling in money from PS3 sales... versus ATI that helped on Wii AND Xbox. I suppose the Apple iPod deal may help a little (Apple is stingy though)

Re:off on a tangent

[mabhatter654]

Via likes their processors too much... it's their president's pet project. Intel has already squeezed ATI out of OEM contracts just like Via on motherboards. Via is a bad match for nVidia though, they're way of thinking is good enough and cheap. nVidia and SUN sound good, if SUN would let somebody else make UltraSparc processors for cheap. That is what nVidia needs to do... combine with a Linux platform they could pull it off, but Sun doesn't really want to open up Sparc either... nVidia is just stuck for now, even though they have top-of-the-line product all the chipmakers and OEMS have carefully worked out how NOT to need them.

Re:off on a tangent

[greazer]

both Xbox and Wii were designed/manufactured by ATI... Correction. ATI didn't design the XBox, nor the Wii. They designed the graphics processors within Xbox and Wii. Furthermore, ATI licensed the IP for these designs, so they don't manufacture the processors. Microsoft and Nintendo handle this themselves.

you see how PS3 is selling? nVidia got the short straw this round. That remains to be seen. The net profit to each graphics company, for each console sale, is dependent on the particular deals they made. Selling more of one console doesn't necessarily mean more profit for that graphics company. Furthermore, consoles typically have a 5 year lifespan. Sales of PS3 over the whole lifetime of the product is what ultimately matters.

Must bundle with GPU

[Macfox]

This won't float unless they bundle it with the next generation GPU. AGEIA haven't been able to get traction with a dedicated card and neither will nVidia, unless a heap of games support it overnight.

Re:Must bundle with GPU

[Kyrubas]

It might be that nVidia doesn't even intend to use the overall PhysX stuff at all, but instead wants to tear it apart for the patents on specific design patents further optimization of their GPUs.

Re:Must bundle with GPU

[mrxak]

That's just it, really. Games need to support it in large enough numbers, and need to do it well enough to make a difference between those without the cards and those that have them. Most people seem to think this is a joke, and the way CPUs are going anyway with extra cores, I think we'd be better off seeing multithreaded games instead of physics cards.

Re:Must bundle with GPU

[Clanked]

Game makers won't require something people don't have, the mass majority won't buy something they don't need. NVIDIA however could throw this on their cards, and have it a big part of The way it's meant to be played. This would really give them something over ATI (other than faster cards :p)

Re:Must bundle with GPU

[Rival]

I see your floating point.

The way I picture things, a Physics Processing unit (PPU?) will end up like FPUs: at first an optional, narrow-use add-on, then integrated on premium products, then more widespread as software vendors feel comfortable relying on it, and finally ubiquitous and practically indispensable.

And then Slashdotters will be able to say, "You kids with your integrated PPUs nowadays -- when I was your age, we had to calculate trajectories and drag coefficients by hand, and we liked it that way!"

Re:Must bundle with GPU

[RelliK]

I always thought that GPU + physics engine would be a perfect combination. Ultimately, the AGEIA card is just a DSP + software driver for calculating physics. A GPU is... also a DSP + software driver for calculating graphics. It wouldn't be too hard to write a driver that does both: some of the pipelines could be allocated to graphics, and some to physics. Might even make a software-configurable to dedicate more/less units to physics.

Re:Must bundle with GPU

[milsoRgen]

The interesting thing about the processor on an AGEIA card is it's similair in design to an IBM Cell processor. Just a fewer number of SPE's...

I can't seem to find the link to the paper that discussed it in detail, if I can find it I'll post it later...

Re:Must bundle with GPU

[RelliK]

Every GPU is similar in design to IBM Cell. It's just a simple but massively parallel DSP with very fast local memory.

Re:Must bundle with GPU

[Rival]

Games are great at motivating the development of better video cards, and to some extent bus speeds, processors and other non-gaming-specific components. This is a good thing, though I have some old-man opinions on how Moore's Law is spoiling many developers.

That being said, I don't believe games drive the adoption of hardware as much as you might be thinking. As a case in point, look at Vista. Ugly and bloated, yes, but perforce nearly everywhere. And the minimum requirements for Aero (which is the one feature your average user is going to jump on -- ooh, it's pretty!) are going to do more to push the next large jump in base video card standards than any given game.

Retailers don't have enough fiscal incentives to stop pushing Vista, even if they do try to gain positive PR by selling Ubuntu or XP on a few low-end models. And if they're pushing Vista, they want to support the pretty interface the public expects. By making hardware-accelerated rendering a practical requirement of the OS, Microsoft has raised the bar of the "minimum acceptable" video card.

Right now we see physics cards as a niche product, barely supported. It has been the same with all technical developments. But if we're heading toward 3D interfaces (which I believe we are,)then physics can only play an increasing roll in such an environment. If that should become the case, then a dedicated processor will be much more valuable then assigning a generic CPU core to try and handle the calculations.

Re:Must bundle with GPU

[milsoRgen]

http://en.wikipedia.org/wiki/Physics_processing_unit#Cell_Processor_vs_PPUs
http://en.wikipedia.org/wiki/Physics_processing_unit#GPUs_vs_PPUs

There are differences. Otherwise Sony wouldn't have wet themselves when they announced Cell technology in the PS3 or Microsoft could of countered their ATI GPU was pretty much the same thing or more powerful or however the market types would of spun it if that was the case

Re:Must bundle with GPU

[mikael]

With the current PC architecture, the CPU has to send data to the Physics card, read the data back, then finally send it down to the GPU. This would have to be done for things like character animation (ragdoll motion), particle systems for visual effects (bouncing off the scenery/characters). Ideally, you would want the Physics processor to have a direct path to the GPU. Then you could avoid two of these steps.

And if nothing else, Nvidia also get a team of engineers who have worked together and have both DSP and current game industry technology experience.

Re:Must bundle with GPU

[644bd346996]

Don't forget that PhysX has software out there, too. It hasn't been doing well against Havok, but it's obviously in NVidia's best interests to promote the use of physics engines in games, seeing as they could provide the hardware acceleration for them. I expect the PhysX engine will soon have the ability to use NVidia GPUs, and it will pushed as a more viable competitor to Havok, especially since Intel cancelled Havok FX.

Re:Must bundle with GPU

[Grave]

And then Slashdotters will be able to say, "You kids with your integrated PPUs nowadays -- when I was your age, we had to calculate trajectories and drag coefficients by hand, and we liked it that way!" But I already say that...

Re:Must bundle with GPU

[mabhatter654]

Intel has already killed that train.. they made GMA950 and X3100 to be "just enough" that Microsoft would certify them for "full" vista effects. Once that happened, gaming on any store-bought PC is pretty much dead under $1,000. Both Microsoft and Intel and the OEMs want to milk the market and charge twice the profit for "gaming" PCs even though the low end PC now is twice as fast as 3 years ago... except for the 5 year old graphics chip!

Re:Must bundle with GPU

[OnanTheBarbarian]

This is true as far as it goes, but not all the results of physics computations can simply be left on the GPU. This is OK for visual effects and improving animation, but if the outcome of a physics computation has an effect on the game world as a whole, then it needs to be sent to the CPU anyhow (as a rough example, suppose a defeated enemy drops his sword, which bounces down a cliff - this is more than just eye candy).

'Jiggle physics' and particle systems, of course, can stay on-GPU.

Re:Must bundle with GPU

[cheier]

The means already exists to use the GPU as a general purpose physics engine. For some reason, this deal doesn't surprise me. NVIDIA gets a team of hardware engineers out of the deal, and IP for PhysX that they can now convert into CUDA. The unfortunate side is that CUDA is an NVIDIA only API for their GeForce 8 series GPUs, thus eliminating any chance of being supported on any of ATIs latest GPUs.

It wouldn't surprise me to see an SLI setup used, but using the PCIe bus specifically to offload physics math to the second GPU nearly exclusively, since through the bus and CUDA, they can also access all of the GPU memory on board the second card.

Re:Must bundle with GPU

[StargateSteve]

what about a hybrid card? One half physics, one half GPU? It would make the whole thing less expensive for consumers, and would deliver some killer physics the ATI couldn't touch.

Re:Must bundle with GPU

[Wesley Felter]

There's nothing stopping you from buying a low-end PC and installing a real GPU. AFAIK, most systems with integrated graphics still have a PCI Express slot so you can upgrade.

I also don't see any gouging going in in gaming PCs. I recently built a $1000 gaming PC and prebuilt models with similar specs were selling for $1100-1200, which is not much of a markup.

Re:Must bundle with GPU

[vision864]

Most gamers who have the coin to drop have been down this road before years back, hmm a dedicated pci game accelerator. where have I see that before???

Nobody wants to go through the 3dfx era again as great as it may have been during its existance. and yes i know its not a graphics card its a physix card. Whatever you want to call it - its a damn strap on game accelerator. next thing you know they'll be putting a cute little logo on games "optimized" oh wait ill shut up now.

Re:Must bundle with GPU

[Anonymous Coward]

That's only true from a *very* large distance. The differences far exceed the similarities. Both chips have FMAC units, sure. But Cell is a moderately-pipelined single threaded RISC design while a GPU is massively-pipelined + hyperthreaded CISC. For instance, the Cell has no exp or sincos instruction, and the RSX can read a bilinearly-filtered texel with zero apparent latency by switching between O(1,000) threads once per cycle.

The memory systems of both are also radically different, with the SPU providing a single 256K area of unmapped memory and the GPU providing various caches at various levels with all manner of data reorganization and duplication going on under the covers.

What the hell is a "DSP" anyway, these days? The term was coined to cover architectures that (a) had a Harvard architecture and (b) could pipeline one FMAC per cycle. ALL architectures do those two things now.

The 56000 is ancient history, my friend, and the modern derived architectures are far, far more complex and varied than you give them credit for.

Ironically, in a few years, your comment will be correct, as GPU design is converging with designs like the Cell. But it ain't there yet, not by a long chalk.

Re:Must bundle with GPU

[Cornelius the Great]

But most users who buy PCs never upgrade their GPUs. Most game developers have two choices: write their software for the lowest denominator and limiting graphics complexity on the PC, or work on console titles where all of the hardware are guaranteed to be equally capable.

PC gaming enthusiasts aren't a very large market compared to console gamers, therefore most studios are likely more focused on developing games for consoles.

This hurts PC gaming, no matter how you look at it.

Re:Must bundle with GPU

[MBGMorden]

There's nothing stopping you from buying a low-end PC and installing a real GPU. AFAIK, most systems with integrated graphics still have a PCI Express slot so you can upgrade. Some do, some don't. Really depends on the model. For those that don't, your upgrade options are limited to PCI cards, or maybe a PCI-Ex1 card if you have those slots. None of the cards that will work on such slots are really going to do much for your gaming options.

Re:Must bundle with GPU

[The_reformant]

With the current PC architecture, the CPU has to send data to the Physics card, read the data back, then finally send it down to the GPU. This would have to be done for things like character animation (ragdoll motion), particle systems for visual effects (bouncing off the scenery/characters). Ideally, you would want the Physics processor to have a direct path to the GPU. Then you could avoid two of these steps.
Thats assuming you run your main loop on either the GPU or physics card which doesnt seem like a sensible idea....but then i dont really have a clue about game programming.

Re:Must bundle with GPU

[Anonymous Coward]

Microsoft could of countered

"could have".

Here comes the bandwagon...

[Tpl2000]

I, for one, welcome our new fairy overlords. I also welcome whoever gets rid of this joke.

The Future of Physics

[calebt3]

the future of physics I am personally hoping that the future of physics leads to warp engines.

Re:The Future of Physics

[Anonymous Coward]

I'd be satisfied with socks that stay up by themselves.

Re:The Future of Physics

[bky1701]

I'll be happy when someone starts to enforce the laws of physics. I mean, do you ever hear about someone in court because they violated the second law of thermal dynamics? Didn't think so. What are we going to do when people stop following them?? It will be lunacy, I tell you!

Re:The Future of Physics

[Antarius]

I'd be satisfied with socks that stay up by themselves.

Or underwear that doesn't crawl up your arse...

Or a bridge from Melbourne to Tasmania...

Or a version of Vista that doesn't SUCK!
(To which the Genie replied: "Is that bridge 4 lanes or 6?")

Re:The Future of Physics

[MrShaggy]

Or a sock that can get it up.. http://www.edthesock.com/

Re:The Future of Physics

[ScriptedReplay]

I'd be satisfied with socks that stay up by themselves.

Apply mud. Liberally.

Re:The Future of Physics

[Hillgiant]

Solution: Don't wear socks.

More Fuel For The Nvidia CPU Fire.

[milsoRgen]

It's long been speculated that nVidia has been developing an x86 processor. I think they great work they did on the 8xxx series cards and the fantastic chipsets they have produced really lend credence to this theory. They could make a very strong processor, especially as it appears we're heading to CPU's with heavy GPU integration. AMD Fusion... It's not so much to put a graphics card in a chip. It's to make a CPU do what GPU's excel at. Very exciting times for hardware enthusiasts indeed!

Re:More Fuel For The Nvidia CPU Fire.

[Anonymous Coward]

yeah, an nVidia CPU would be great! if you needed to heat a whole house. going by their GPUs, it'd consume 3x the wattage of a comparable intel/amd and put out 4x the heat.

Re:More Fuel For The Nvidia CPU Fire.

[webmaster404]

I don't really think that all this will be better in the long run. While faster GPUs and better cards mean faster games, with all the DRM that Vista has it makes them more expensive and have poorer performance. Linux lacks in games to really test the cards out and getting drivers for ATI/Nvidia is a pain to say the least, and OS X really doesn't support non-apple internal hardware very well so that's not a test. Technology wise in the hardware department we are making leaps and bounds every day, however with the lack of a decent OS to test the new cards on, their true potential will be lost in DRM/Vista/Driver issues.

Re:More Fuel For The Nvidia CPU Fire.

[milsoRgen]

Just look at Intel's rather quick turn around from the P4 to the Core architecture. They were headed down the same road GPU makers are going, yet reversed course. Sure it's mostly thanks to the Israeli development team that produced the Pentium M. Which was in turn based the Pentium 3. The fact of the matter is nVidia has shown time and time again they can make a killer product. I believe they could make a highly efficient CPU with performance to watt ratios well inline with current products. If not even better.

But on another note... The heat issue with GPUs really does need to be resolved. I'm using a x1800 XT ATI card... And I've come pretty close to 100C at times... I'm not quite sure how current gen cards are doing in this area, but I doubt it's been anything like the P4 > Core turn around.

Re:More Fuel For The Nvidia CPU Fire.

[Anonymous Coward]

it'd consume 3x the wattage of a comparable intel/amd and put out 4x the heat.

In this house, we obey the laws of thermodynamics!

Re:More Fuel For The Nvidia CPU Fire.

[milsoRgen]

Yeah, I agree... Unfortunately.

I don't like being tied into Vista for the latest hardware support when it comes to gaming. Microsoft developers have gone on record as saying there is no reason DirectX 10/10.1 couldn't be used with XP. They just wanted it to be used as a 'dividing point' in which to ditch all the old tech... Something like that. It was in a Maximum PC a while back... Makes ya wonder what happened to OpenGL in the PC gaming market.

But to veer a little, I will say I found that the driver issues were a bit overblown. At least for me, using the RTM edition of Vista a while back. And once I disabled things like UAC and reverted the GUI to 'classic', I really wasn't disappointed in the performance, nor was I impressed. I think the real problem with Vista is that, yes it does consume more resources than is needed or called for but also perception is reality. And the perception is Vista is a pile... I agree to an extent but not as much as some.

Re:More Fuel For The Nvidia CPU Fire.

[MadnessASAP]

I don't want to insult you by pointing out something that may seem obvious but the reason the GPU is getting up to 100C is probably because the fan or heatsink has stopped working. Try checking the thermal pase and blowing out any dust that has built up. The only time I ever say a chip get up to a 100C was when an "A+" certified tech (who is also my uncle) unplugged the CPU fan while installing a stick of ram in my cousins computer.

Re:More Fuel For The Nvidia CPU Fire.

[DAldredge]

What issues does Vista have with DRM when playing non-DRMed content?

Re:More Fuel For The Nvidia CPU Fire.

[khellendros1984]

The drivers still do hardware tampering check....the "tilt bit". If the driver detects anything screwy with the hardware (voltage fluctuations, etc), it resets the machine. Even when running non-protected content, this is being checked, which reduces performance and makes it possible that the machine will reboot at arbitrary times. http://talkback.zdnet.com/5208-12558-0.html?forumID=1&threadID=28793&messageID=537882&start=26

Re:More Fuel For The Nvidia CPU Fire.

[The Analog Kid]

Makes ya wonder what happened to OpenGL in the PC gaming market.

Maybe if OpenGL development was run as a dictatorship rather than an oligarchy, it would catch on. Right now it's lagging behind DirectX featurewise, because development by committee is slow.

Re:More Fuel For The Nvidia CPU Fire.

[mikael]

I had the same problem with a latop - turns out the intake vent (on the underside of the laptop!) for the cooling fan was clogged with dust - using the narrow tip of vacuum cleaner cleaned out all that junk. No more "CPU is running in modulated mode" messages.

Re:More Fuel For The Nvidia CPU Fire.

[westlake]

While faster GPUs and better cards mean faster games, with all the DRM that Vista has it makes them more expensive and have poorer performance

You do not have to max out your credit card to get good performance in a DX10 card:

The Radeon 3850 brings us something we've been begging for ever since the DirectX 10 cards were introduced: a sub-$200 card with performance comparable to the high-end products. The Radeon 3850 delivers Geforce 8800 GTS 320mb performance for $100+ less. If you're looking to get the best possible performance for the dollar, this card hits the sweet spot. Best Gaming Graphics Cards: February 2008

Vista can play protected HD content at full HD resolution. The PS3 can do this. OSX can do this. The set top box with an embedded Linux OS can do this. No HD means no mass market sales.

Re:More Fuel For The Nvidia CPU Fire.

[RiotingPacifist]

Inovation died :( theres no real inovation in games, they all trundel along in the same direction. better graphics, better physics, better AI. Never anything more than better tho. It would be nice if a they put some effort into new areas, and saw what happened. AS a result they're no point in selling a game to everybody just trundle it out to the masses.

Its a real shame that game development doesnt lend itself to OSS, there are so many areas where abit of Cross genre thinking would produce groundbreaking games (like the multiplayer game where you manipulate the levels (it was around before Garys-mod and allows alot more). I really think AI is the way games need to go, as neural nets mature ( say 3-5 years ) the AI could generate content and try to make every play different, for a standerd game developer this would be alot of work, but for OSS it would be half done already.

Its a shame about openGL, i think the standards -> code/hardware way of working is inherently flawed, it would be much better if it was code(a) -> standards -> hardware -> code(f). where the code(a) would be developed with new features and code(f) would be have whatever the standards endorse. The development model would be quite nice too as code(a) would include people working on a whole host of different functions ( designer stuff, game stuff, embedded stuff, etc) which even if rejected by the standards body could be implemented in specific hardware.

Re:More Fuel For The Nvidia CPU Fire.

[Man Eating Duck]

...probably because the fan or heatsink has stopped working.

That's true in my experience as well. I bought a second-hand GPU some time ago, and experienced severe instability. I fired up a hardware monitor, and sure enough, it was chugging along at 100-108 degrees Celsius under load. I was quite astounded that it didn't burn out.

I opened the case, and the fan and air ducts on the card was completely clogged. The fan spun, but moved no air at all. I removed a solid block of lint from the ducts after disassembling the card, problem solved. The temperature sensors weren't broken either, as I now get around 40-50 degrees on the GPU under load.

Incidentally, even if you don't have such serious problems, cleaning the various fans in your case for dust and lint will make your computer a lot more quiet, both from slower fan speed due to more effective fans and lack of friction between the fan blades and the air. Recommended at the price :)

Re:More Fuel For The Nvidia CPU Fire.

[Anonymous Coward]

Based on your first paragraph, I'd have to guess you either haven't played any games made since 2000, or you're just one of those useless people who thinks everything old is better because it's old.

Re:More Fuel For The Nvidia CPU Fire.

[Anonymous Coward]

microsoft killed opengl as a possible gaming platform -- a lot of people forget, but some years back a project was announced which would give direct3d and opengl unified driver cores at an operating system level (in windows). (yes, i know that vendors who support both opengl and d3d do this already nowadays, but this is before such inventions)

this was a joint project between sgi and microsoft (although i forget the name now) - it was not long after microsoft had pushed opengl to one side with their direct3d by stating that d3d was more designed for games, but that opengl was not suited to games and was only suited to scientific visualisation apps, etc.

it was an amazing point in history, because for a while at least, microsoft looked like they were going to offer equal support to both their own d3d and the more open-standards-based opengl. for a while it all looked good.

of course, a bit further down the line, and the project went to pieces - i don't recall what happened, but suddenly microsoft's support for an integrated hardware supported opengl core got dumped, the deal with sgi got cancelled.

instead, at the time we got a new d3d from microsoft, and limited support for opengl from a few vendors who thought it worth supporting (thanks nvidia!) - but no thanks to microsoft really for that.

many people don't remember the above incidents very well, or at all -- but it's all just another case of microsoft messing with 'standards' and refusing to play nicely.

i am sorry, but i no longer recall what this sgi-microsoft joint project was called.

Re:More Fuel For The Nvidia CPU Fire.

[JasterBobaMereel]

Fahrenheit project - was going to be OpenGL meets Direct3D, Microsoft backed out and it collapsed, and then they stopped supporting anything but the very basic OpenGL core system on Windows (and almost completely on Vista)

OpenGL is supported by all the consoles (except the XBox of course) and all major operating systems (Except properly by Windows)

It is still alive well and going forward and is used in Graphics systems other than Games (Direct3D was only for games)

Re:More Fuel For The Nvidia CPU Fire.

[Cornelius the Great]

You're referring to Farenheit. SGI put a lot of effort into development, but MS decided they had nothing to gain from the partnership and continued work on DirectX chose to ignore the deal they made with SGI.

Re:More Fuel For The Nvidia CPU Fire.

[BlackSnake112]

Though the graphics card you are using does everything you want it do. The newer cards from both ATI and Nvidia are lower in power and heat output. Well, unless you over clock them.

Re:More Fuel For The Nvidia CPU Fire.

[RiotingPacifist]

Actually i played games from about 2002-2006/7 but there are very few news ideas its just a constant progression, i mean HL2 was just like HL1 but with modern graphics and physics. The only place ive seen innovation has been in 3rd oarty mods but they never get much attention anyway

Re:More Fuel For The Nvidia CPU Fire.

[milsoRgen]

I'm not sure how I got modded flamebait, I'm simply trying to point out some things.
1. GPU's tend to dominate when doing certain calculations. Most notably floating point intensive applications . See: Folding@home. Therefor with the current stated intentions of both Intel and AMD to incorporate GPU type capabilities into their general purpose processors nVidia needs to further innovate above and beyond their current level.
2. By incorporating an already tried and true, robust hardware based physics model into their GPU's nVidia would have a tremendous advantage compared to Intel and AMD (unless the AMD Fusion concept comes along nicely).
3. The only thing potentially holding nVidia back would be basic integer performance, doesn't seem to far fetched for them to make something that would be in line with current processors there, if not exceeding current performance, although as people have pointed out heat is an issue, so I would expect some or all of the chip to be scaled back to a degree. But I also wonder if licenses might be in issue. However I do not believe so as nVidia has been licensed to develop chipsets for both AMD and Intel processors... Though licensing issues aren't really my forte...
4. All that being said, I think now more then ever, nVidia is in a very good position to release an x86 processor.

I wonder how this will affect AMD's GPU offerings

[Scareduck]

I don't pay close attention to the GPU market in general, though lately I've been interested in a few numerical modeling projects that could benefit from high-performance computing. The AMD Firestream 9170 is supposed to be released in the first quarter of this year, with a peak speed of 500 GFLOPS, most likely single-precision, but the beauty part is that it should also support double-precision, the numeric standard for most computational modeling. NVidia's option in this space is the Tesla C870; I wonder whether this move to purchase another GPU line will divert resources away from their number-crunching-first GPUs.

Not Good

[sexconker]

Intel has Havok, Nvidia has Ageia, AMD/ATI (DAAMIT) has nothing.

So developers will have to make 3 versions of the game, then?

Can't wait for DirectX 11(tm) Now with Fizziks Power (tm).

Re:Not Good

[Loadmaster]

Oh noes! There's no hope for Duke Nukem Forever if this happens! They can't complete one game let alone three.

At least Satan can put away his parka for another year, and we can rest assured the sun is not going red giant anytime soon.

Re:Not Good

[mrchaotica]

Can't wait for DirectX 11(tm) Now with Fizziks Power (tm).

You say that sarcastically, but I actually would like to see something like "OpenPL" (Open Physics Language).

Re:Not Good

[DeKO]

That's pretty much unfeasible. Every game needs a different physics simulation. Rigid bodies, ropes, soft bodies, particles, cloth, and so on; each requires a very different strategy. And there are many special cases where you can customize the algorithms for your specific simulation; using a more general algorithm when a specialized one is possible is less efficient.

And this doesn't even get into the details about strategy; continuous vs fixed time steps, different orders of integration, collision detection and so on. Each has its own quirks; and Nintendo is proving us all the time that you can create superb games using almost no physics.

Re:Not Good

[Cornelius the Great]

Well, there's already Open Dynamics Engine.

Whither Nvidia/PhysX?

[Crypto Gnome]

1. Purchase Aegia
2. Continue selling dedicated Physics addon-cards
3. Integrate PPU onto Graphics Cards
4. (somewhere along the line, get full Microsoft Direct-Something endorsement/support of dedicated physics processing)
   • possibly by licensing to AMD "PPU included on Graphics Card" rights, thusly invoking the power of Least Common Denominator
5. Integrate PPU circuitry/logics into GPU (making it faster/more efficient/cheaper than equivalent solution licensed to AMD)
6. ?? Profit ??

In the end, for this to *really* succeed, it needs to be a "Least Common Denominator" factor. So it *requires* full support by Microsoft and Direct-X (them being The Big Factors in the games industry). And in order to get full support from The Windows Monopolist, you'll probably (not absolutely necessary, mut it'd make it much easier to convince Microsoft) need to enable AMD/ATI to leverage this technology, to some degree.

Remember folks, Nvidia don't need to *kill* AMD/ATI, they only need to stay one or two generations ahead of them in technology. So they *could* license them "last years tech" for use on their cards, to make "Least Common Denominator" not a factor which excludes their latest-get tech implementations.

Re:Whither Nvidia/PhysX?

[Antarius]

Nvidia don't need to *kill* AMD/ATI, they only need to stay one or two generations ahead of them in technology. So they *could* license them "last years tech" for use on their cards, to make "Least Common Denominator" not a factor which excludes their latest-get tech implementations.

I wish I still had Mod-Points, 'cos that deserves a +1, Insightful!

Yes, people seem to forget that business doesn't have to be ruthless. Sure, you can take that path and it has been proven to be effective by people in many industries, including IT. Punctuating your sentences with chairs can also help emphasise a point.

Many successful large companies quickly learn that the "Us vs Them" mentality isn't always necessary - and licensing IP or standards in this fashion can be quite lucrative! (Oh no... I made a positive reference on Slashdot that valid IP & standards being alright to license for profit... There goes my Karma!*)

Intel's licensing of its' SSE extensions to their competitors is a good example of how a standard can be strengthened and made more effective by 'working with' their competitors, as was AMD's licensing of x86-64 to Transmeta.

Of course, this is NVIDIA we're talking about. The likelihood of them licensing it, even for profit, is about as high as Microsoft donating millions (of dollars, not bugs) to the WINE project...

*For the FRZs, I am against Patent Trolls, but for a company/individual's right to profit from a defined standard if another company wants to benefit from their R&D rather than re-invent the wheel! This is, of course, completely different to Joe Scumbag getting a Patent for some-general-nose-picking-device (idea only, no intention to develop) and then extorting any companies that then try to develop a real nose-picking-device. That would be "Just Plain Wrong(tm)"

So you see, I'm a good sycophAnt... I hate Darl McBride too! Don't take it out on my posts, please!

Re:Whither Nvidia/PhysX?

[fyrewulff]

Good luck with that then. nVidia screwed Microsoft over on the Xbox, and MS has been making them feel it ever since.

Interesting news.

[Besna]

The computing industry is seeing a dramatic shift towards single-package parallelism. Yet again, the x86 architecture largely holds back the CPU from becoming more all-purpose and doing GPU and PPU activities. There are actual engineering reasons you can't have a truly general-purpose ASIC (you can with an FPGA, but that would be too slow for the purpose). The GPU and PPU is where the interesting stuff is. They can actually write new macroarchitecture! They can design on-chip parallelism with far greater complexity without the need for a backwards-compatible architecture.

The exciting aspect to this acquisition is the stronger fusion of two companies that have the ability to harness processing power without historical limitations. ATI/AMD really didn't have this, with AMD stuck with x86. Something like Cell is interesting in this space. However, it lacks flexibility in matching up the main core with the secondary cores. Why bring in PowerPC, for that matter?

This will lead to great things. It is fun again to follow computer architecture.

Why do we need physics cards?

[Cathoderoytube]

So, I'm assuming I'm not getting all the physics simulation quality I can get out of my games? The whole deal with the bridges collapsing in real time and all sorts of junk bouncing around isn't the ultimate physics experience? Is there... Another level of ragdoll I'm not experiencing? Is there some dynamic to a flaming barrel rolling down a hill my computer can't handle?! Or.. Or.. Is it Nvidia making one of its patented cash grabs?! Considering all the physics simulations in games to date have been done on the processor with no performance hit (Have you played the last level in Half-Life Episode 2?) I'm finding the notion of dedicated physics card fairly stupid. But that's just me.

Re:Why do we need physics cards?

[dbIII]

Consider something like WoW - lots of pretty pictures but you can run through people or monsters. Collision dectection requires a bit of effort when there are a lot of objects.

Re:Why do we need physics cards?

[maglor_83]

Thats not why there is no collision detection in WoW. You already know what objects are nearby - the server isn't sending you location data for every person currently logged on. There is no collision detection because otherwise a whole bunch of jerks are going to line up in a row to stop people from getting to where they want to go.

Re:Why do we need physics cards?

[Grave]

That and the fact that to have that sort of real time physical accuracy, you must also have enormous amounts of data going back and forth in real time, which would make the game laggier than it already is in high-pop areas.

Re:Why do we need physics cards?

[NeMon'ess]

With physics acceleration, the little things that don't feel real could be done.

Running through grass could cause it to deform and brush the character, and some of it gets stepped on and stays bent down. Or in sports games, each limb could have a better defined clipping box and rules for how it can flex.

Then when two players collide going for a ball, they hit more realistically and don't clip through each other. Especially on the slow motion replays it would look nice.

Or in a racing game, when cars crash, they could really crash. Imagine bodywork deforming and "real" parts going flying, instead of only a flash of sparks.

Also, it would be cool for grenades and other explosives to properly damage the room and buildings in games that want realism. Walls that crumble into rubble. Tables that break into chunks and splinters. Ceilings that collapse when the supports are destroyed or weakened too much.

Then outside, no more indestructible walls. When I ram a truck or tank into an unreinforced building, something actually happens. As in the vehicle crashes through the wall, or continues through the building with momentum.

Re:Why do we need physics cards?

[Charcharodon]

Actually the thing you are missing is the bridge pieces bending before it collapses, the barrel being dented as it rolls down the hill, or the rag doll limbs breaking or being ripped off with the proper application of force. Those things cannot be done in real time on a CPU.

Unfortunately most of those things are only avialable in demos atm. UT3 has a couple of special maps that do some neat stuff, but then you start running into problems with the video card trying to keep up with the 100 or so bricks that just came crashing down from the wall you just demolished.

In the main game the only physx I noticed was the cloth simulation of the flags, and the main characters outfit, of course you don't exactly have a whole lot of time to take this all in since everyone is trying to kill you.

Ageia is not just hardware physics the software they make does a pretty good job. The vehicles in UT3 are some of the best I've every seen. I started laughing during one game because I managed to get a small vehicle wedged under my tank and kept going, dragging it along, verses instantly getting stuck while the CPU sits there trying to figure out the clipping and collosion detection.

The premium idea in my book would be for Nvidia to integrate the function into their video cards, but keep it dormant, so that it is only used as a video card at the time, and then when you upgrade your video card, the new one takes over the video and the old one moves over a slot and becomes a dedicated PPU in the second SLI slot.

Re:Why do we need physics cards?

[TheThiefMaster]

Tables that break into chunks and splinters. Almost every game nowadays has breakable objects. The chunks are generally pre-determined, not generated based on where you hit, but it works.

Or in a racing game, when cars crash, they could really crash. Imagine bodywork deforming and "real" parts going flying, instead of only a flash of sparks. If you want something like that that's out now, check out FlatOut and FlatOut 2 on Steam. The cars are fully physics simulated, and a lot of the track is destructible.
But still, the reason cars can't be damages is normally a licensing issue. "You can use our cars, but they can't be damaged". FlatOut (2) has no licensed vehicles.

Destructible buildings though, that WOULD be cool.

Re:Why do we need physics cards?

[MrAngryForNoReason]

Running through grass could cause it to deform and brush the character

Crysis has this, it isn't entirely accurate but the grass and tree branches move as you push through them.

grenades and other explosives to properly damage the room ....Tables that break into chunks and splinters.

Most games from the last few years have destructable objects, as said elsewhere the 'pieces' are pre-determined but the effect is good enough.

Ceilings that collapse when the supports are destroyed or weakened too much......When I ram a truck or tank into an unreinforced building, something actually happens.

Again Crysis has done this, if you drive a truck into the weaker buildings they collapse into sheets of corrugated steel and wall panels.

Physics simulation has come a long way without stand alone physics chips.

Re:Why do we need physics cards?

[grumbel]

Almost every game nowadays has breakable objects. The chunks are generally pre-determined, not generated based on where you hit, but it works. It "works" in the sense that it looks rather ridiculous. You might not notice it in a fast action game, because you are in the next room before the chunks and splinters hit the ground, but for actual interaction with the game world the current stuff is just way to simple. One of the easiest examples is rag-doll animation, which looks nowhere near what a real human would look, its not even close and the reason is simply that accurate simulation would need a bunch more processing power then is available.

But still, the reason cars can't be damages is normally a licensing issue. Licensing is an excuse, not the reason. PC games had car damage since basically forever, it is just Grand Turismo that is waving the "licensor" flag as excuse.

The problem with physics in games currently is that they are mostly for show, not substance. Which is sad, since real physics modeling can increase the immersion and fun in a game a lot. But shooting a grenade into a wall today is still the same thing as doing it in Doom1, which is sad, since Doom1 is 15 years old already.

Today we are simply still at the very beginning of physic development in games. In games today only a very tiny part of the world even reacts to physics and this is something I expect to change in the coming years and having physics acceleration at hand will certainly help.

Re:Why do we need physics cards?

[TheThiefMaster]

The biggest problem in FPS game physics at the moment is a ridiculously trivial one:
When the player moves forward he pushes whatever is under him forward.

I'm not kidding, try standing on something and moving. In reality there's no real problem, as long as the centre of gravity of you+object is under the object's base. If it's light enough, you might kick it out behind you. In a game, 99% of the time you will kick the object under you forwards and out from under you.

Try standing on a barrel in Half-Life 2 for example. The game even gives you a puzzle at the start where you have to stack crates to get through a window, where you can notice this problem.

Re:Why do we need physics cards?

[Cathoderoytube]

I don't think you actually understand what would fall under the physics category. Grass can be deformed with driven keys, no need for fancy physics cards there. Players crashing into each other would be controlled by animation. I don't know where people get the notion animation can happen on its own without people sitting down and keying in everything beforehand.
As for the racing game. Pick up a copy of Dirt and then try shooting your mouth off about non realistic crashes.

Overall it looks to me like you stopped playing games maybe four years ago. We have breakable tables and damagable rooms in games now. It's all very interesting and it doesn't need a physics card to do.

Re:Why do we need physics cards?

[Justus]

Ah, yes, that makes sense. Let's use Crysis as an example for why dedicated physics processors wouldn't have an impact on performance. After all, high-end systems running Crysis on higher quality settings (which presumably impacts the physics simulation) get an average of 30fps. No room for improvement there, right?

Re:Why do we need physics cards?

[Charcharodon]

I wouldn't get too excited about Valve, last I read they are skipping over support for 2 and 3 core CPU's and going straight to 4 for all their future goodness, so it's going to be a while before that hits the mainstream.

I still like the idea of a PPU, but it would be nice if it were intergrated straight into the GPU or the CPU, or a chip on the Mobo, just like so many other dedicated devices that have come before it.

I just wish Nvidia and ATI and 3rd party software and start seeing some neat GPU based things that would make things other than games scream.

Re:Why do we need physics cards?

[Cathoderoytube]

That's a bad example. You don't want collision detection in a game like Warcraft. There are so many people running around it'd turn the game into an ordeal just getting to the auction house in Ironforge.

Re:Why do we need physics cards?

[Cathoderoytube]

Real time physics only really appeals to folks who for some reason think that everything done in real time is the way to go because it's real or some nonsense like that. I haven't read any gaming magazines in a while so I'm not sure what the lingo is.

Gamers need to believe they're getting some real time bloody experience so the thought of real time chunks of debris falling gently on the real time terrain is appealing to them. It's very good marketing since gamers fall for that sort of horse shit. I mean if I had it my way I'd rather see bloody raytracing implemented into games, not that I'd want that, but it'd be a shade more useful than real time junk that's handled perfectly well on the processor.

Take your great barrel scenario for example. It's quite simple to model a dented barrel and change the visibility or put some driven keys on a barrel to make it dent when it gets knocked down. You don't need to apply real time physics to it, in order to make that happen.

You apparently have no idea what handles the processing of physics simulations. The video card would only be responsible for processing the polygons which make up the falling blocks. In a case like that you need a faster video card. A physics card in a case like that isn't gonna make much difference.

As for your anecdote about UT3, you can do the same thing in other games without physics cards. Once again you have absolutely no idea what physics cards do, why are you arguing for them?

And I agree with you the ideal situation would be Nvidia putting the physics processing on their stupid cards, just to avoid people having to go out and buy two seperate cards. At the same time ATI doesn't have the technology so what does that mean for them?
I would assume Nvidia will be pressing game developers to implement their physics technology just so they could force more people to buy their cards. I doubt ATI would have any of that.

Re:Why do we need physics cards?

[Charcharodon]

You seem to be rambling about something, but I'm not sure what your point is.

The video is the video and the physics is the physics. Currently right now all the physics is handled by the CPU or by the PPU if you have it and software supports it.

I think you are going on about how "game physics" is not the same as real physics, which is pretty obvious to anyone that's been following the tech. Fully blown real material physics simulations couldn't run in real time on any machine currently known to man and probably not anything that will exist in the next 20 years to be usefull.

I could give a shit less where the game physics runs, as long as it's being developed. In an extra card, then so be it. Integrated into something else, cool. Running on an unused core on a multi-core CPU, bonus.

I could care less if ATI get's left out in the cold. They've been going on for years how physics could run on their video cards with nothing to show for it. AMD has been promising vaporware for a while too. Maybe Nvidia snapping up Ageia will light a fire under their ass and get them to move some hardware and software out the door.

Re:Why do we need physics cards?

[Anonymous Coward]

The biggest problem in FPS game physics at the moment is a ridiculously trivial one: When the player moves forward he pushes whatever is under him forward.

Isn't that a programmer error, i.e. not understanding conservation of momentum and solving the equations wrong, getting the sign wrong?

Re:Why do we need physics cards?

[Cathoderoytube]

Thanks for the heads up idiot.

Re:Why do we need physics cards?

[NeMon'ess]

I don't think the current effects look good enough. What about when walking through ferns or small bushes? Touching a few inches of the plant or a foot of it means it should bend less or more.

Animating player collisions is why they don't look good enough. In a football game, the angles of collision are always going to be different. The arm positions are going to be different as a defender and receiver reach for the ball. When the people collide ought to look different each time.

Just saw the trailer for Dirt. Pretty impressive. I could say it's good enough for that game, where the race must go on after a crash. How about in a game where there's lots of accessory cars? When I run over them in a tank, the hood, passenger compartment, or trunk is going to crush differently depending on if I run over it from the front, side, or diagonal. Or how about if I crash a car into a telephone pole or instead of crashing dead-on into a building, it's an offset crash?

I don't know where people get the notion animation can happen on its own without people sitting down and keying in everything beforehand.

People get that cloth can be simulated. So it's easy to imagine that a body panel is made up of polygons, because it is. Give the material ratings of how easily it bends and deforms, then calculate for the length of the side of the triangles and the angle of attack. If there's too many polygons, then give the object a low-poly mesh the deform instead.

I've also seen walls explode in animation. It looks like crap because the wall is a solid panel, or maybe it breaks into two or so pieces. Except it doesn't match the force of energy applied to it. Add 2x4s every 18 inches, or make it 36 for easier rendering. Animate the drywall near the explosion disintegrating, but farther from the blast there should be small chunks, and ragged chunks still hanging on the structure left standing.

Something else I thought of, couldn't physics acceleration be used to do water right? I've seen Crysis and it's water looks half way real and half way fake. Why is it that Wave Race 64 for the Nintendo 64 could actually animate the surface of the water, but basically everything since still uses a flat plane? A person or vehicle traveling through should generate more than just a ripple animation or froth of the flat surface and for larger vehicles some particle splashes. The surface should deform, which physics can do nicely.

Also, when something exits water, it doesn't look wet. Or on beaches, I've seen wave animations on the sand, but the sand doesn't look wet, because the pixel shaders aren't making it any shinier. The most that happens is the texture on the beach appears to show the edge of the waves going up the sand.

Re:Why do we need physics cards?

[NeMon'ess]

I would assume Nvidia will be pressing game developers to implement their physics technology just so they could force more people to buy their cards. I doubt ATI would have any of that.

Back ten years ago, when games were ten times cheaper to make, 3dfx had games that only worked on their cards.

Then around 2000, games ran on both the Riva TNT 2 Ultra and the GeForce 256. Except the second card had hardware T&L and the Riva did not. Games with the effects looked better on the GeForce.

Once half the cards on the market (nVidia) support physics acceleration, there will be games that look awesome with the effects. Cards without the acceleration won't deliver as pretty an experience. Consequently gamers will buy more nVidia cards.

Re:Why do we need physics cards?

[banana fiend]

That's because most games programmers implement their character as a solid object (until they ragdoll due to collision or death). The object is moved by pushing from behind rather than the "feet" pulling forward. It's not a problem with the physics so much as a problem with optimised implementation in games.

A physics card is just dual-core for the idiot

[idonthack]

With dual-core coming standard now on all new PCs, and multi-core rapidly approaching, physics cards are done for. Graphics cards are still a good idea because the kind of calculations they do can be heavily hardware-optimized in a way that general purpose CPUs are not, but physics cards don't do anything a second (or fourth) full speed CPU isn't capable of doing better and faster.

Re:A physics card is just dual-core for the idiot

[gbelteshazzar]

this is the key, quad core cpu's are around the corner, one core for physics. sure a cpu core may not be great at physics (well, not as good as a dedicated chip) but consumers won't see much difference. may use one of the cores of a quad core gpu. anyway, the dedicated physics card is dead in the water. pull the patents out and plug them into the gpu.

Re:A physics card is just dual-core for the idiot

[XaXXon]

Shoot. With all those available cores, let's move everything back to the CPU. Get rid of graphics cards.

Oh wait. General purpose CPUs aren't very good at certain types of workloads.

Re:A physics card is just dual-core for the idiot

[Tolkien]

You're missing the point.

The whole point of a physics card is to move the calculations away from the CPU (which is so generalized it can't be optimized better than a hardware implementation). Having a card dedicated to processing physics simulations means it gets 100% of the PPU's attention instead of 10% of CPU1's attention, 11% of CPU2's attention, 5% of CPU3's attention and 13% of CPU4's attention (This is after all a PPU, not a CPU). Not only that, the PPU and the hardware on which it is set is optimized for pure physics calculations, not graphics, not playing music or mailing spam: physics. This, a CPU is not. This leaves more CPU time available for those irritating background processes that kick in at all the worse possible times during multiplayer games, meaning those impact your game's performance that much less.

It's the exact same principle as a graphics card. Yes current non-physics hardware can simulate half-decent physics, though in practice, ragdoll physics et al get repetitive very quickly, still better than prefab animations though. Good luck successfully reproducing REAL physics without specialized dedicated hardware.

Re:A physics card is just dual-core for the idiot

[Tolkien]

I didn't think of this until after I posted, but how do you think graphics cards came about? They started off integrated with motherboards too, then it was discovered that dedicated hardware can perform MUCH better relatively cheaply. Same deal with sound cards, ditto network cards, what with the KillerNIC now. It's pure logic really, specialization leads to better performance.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

So you integrate a GPU and PPU into the CPU much like your FPU. Don't you remember the 386/387 and earlier? When the 486 was released, the biggest change was it had an integrated FPU.

Re:A physics card is just dual-core for the idiot

[Webmonger]

I didn't think of this until after I posted, but how do you think graphics cards came about? How did graphics cards come about? Wow, you must be young. Let me tell you.
Sometimes, when a mainframe and a television set love each other very much...

Naw, man. At least in terms of the PC architecture, graphics hardware has been available in expansion card form since the original IBM PC offered your choice of MDA (text only) and CGA (limited four-color) cards.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

Umm, no. Discrete graphics cards were always required until relatively recent years when they started integrating them on to motherboards. The same for sound cards and network adapters.

I remember the full length CGA for my solid steel cased Kaypro 8086, the Paradise EGA in my 286, the Trident VGA adapter in my 386, the Cirrus Logic SVGA in my 486, the Matrox Millennium II in my Pentium, the dual Voodoo 2's in my K6-2, the Voodoo 3 in my Pentium 3 and the Geforce 256 in my Athlon. It wasn't until after that when integrated video become standard fare for motherboards.

You must be either incredibly young or incredibly naive to not know this.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

Did you even read the post you replied to?

Re:A physics card is just dual-core for the idiot

[idonthack]

I get the point of a PPU. But at this point brute force is cheaper and easier. CPUs do well enough at processing physics, and they are fast enough and cheap enough that it's smarter to run respectable physics code there instead of spending an extra $100 to $200 on an expansion card that provides almost zero gameplay enhancement. PPUs may have been a good idea ten or fifteen years ago when CPUs were slower, but game physics engines haven't improved significantly since 2003. Faster processors are available cheaper every day, and they only have to run physics code that ran well on even mediocre CPUs five years ago.

Improving game physics is now simply a matter of turning everything in the game world from a static object into a physics object, which will not even provide a significant gameplay improvement. Games like Crysis have nearly done it and they weren't even designed to run with a PPU.

Re:A physics card is just dual-core for the idiot

[RiotingPacifist]

Sure for a 1st build $100-$200 may seam like a waste but for my 5 year old computer a phyics card is going to save me buying a whole new proccessor + mobo + gfx card , and then when i do upgrade i can keep the advantage of the physics card. It would also make integrating physics into everyday computing alot nicer, compiz isnt so nice on a buzy integrated chip as it steals so much CPU, but with a GPU it barely affects what your doing.

Re:A physics card is just dual-core for the idiot

[sssssss27]

I don't think you are thinking grand enough. I remember the days where you didn't need a dedicated graphics card to play games and I'm only 21. You really didn't get improved game play per say but it did look a heck of a lot better. A dedicated physics processor though has the potential to vastly improve game play and realism.

Imagine instead of designers creating models of buildings they actually built them. That is a brick building had individual bricks all stacked on each other. Whenever you hit it with an explosive it would actually crumble like a real building or burn like a real building. That is a lot of calculations which a general CPU isn't the best at.

The thing is not enough people have PPUs in their computers so you can't include it into core game play yet. Hopefully nVidia acquiring Ageia will allow them to start bundling it with their GPUs or even better yet offer it embedded on their motherboards. While graphics are easily scaled, game play elements are not. I wouldn't be surprised if you see PPUs being crucial to the game on consoles before PCs.

Re:A physics card is just dual-core for the idiot

[Tolkien]

I'm 25. The first computer I toyed with was an Apple IIe. :) At the time though I was 8ish, so I got to enjoy the games in those nice big 5¼ floppies, heh. That's where my knowledge begins. :) Programming came in grade 5-6 and after though.

My point remains (and this is where I rephrase to remain what I figure is correct: ), the reason they get integrated (aside from early releases and prototypes and such) is because the technology in question has gotten mature enough and/or plateaued in advancement, at which point it takes another breakthrough before it becomes popular to split that component off into its own hardware piece again. Then there's the fact that it's good for business, etc etc etc.

I'm probably mildly (or wildly) inaccurate in some respect because there are 1001 interpretations of every aspect and issue I touched on, so feel free to correct me or add to what I've said. :)

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

486DX had an FPU integrated, the SX was FPUless and required either an 80487 chip, or a replacement 486DX chip in a socket provided on the motherboard.

Thankfully I skipped that whole debacle and had a DX from the get-go, but there were mobos set up for one or the other.

Re:A physics card is just dual-core for the idiot

[UncleFluffy]

The classic description of this process is the Wheel Of Reincarnation.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

The real reason that things became integrated was cost. It was simply cheaper for the manufacturer to slap a cut-down display chip on a motherboard (usually sharing video memory with system memory) than it was to provide a full card. But you will also note that discrete cards NEVER disappeared during this time. Integrated became an easy and cheap solution to sell PCs to the masses and businesses, where performance and upgradability don't matter much. This is not to say that integrated components cannot be as fast or faster (look at the gaming laptops) than discrete components, but if you look at the motivation for it (cost), it wouldn't make much sense to integrate a high end part with supporting logic circuits.

One product comes to mind in all of this. The original Sony Playstation. By all accounts it should have been far weaker than the PCs of the era, however it consistently beat them in terms of graphics ability. Why is that? Because its lowly 33Mhz MIPS CPU had integrated components such as the GTE and MDEC, which were capable of providing services to the CPU much faster than any external bus driven part could.

My hope is that AMD or Intel begin to replace the internal FPUs of processors with GPUs.

Re:A physics card is just dual-core for the idiot

[Loibisch]

I disagree. Physics processing needs a lot of vector math, which your run-of-the-mill floating-point-based CPU is not very good at (even with SSE optimizations). So what you are saying is you prefer to let your 4th CPU core do some very specialized work that a card with about a quarter of the transistors of your CPU can do very efficiently at probably the same speed.

People were mocking 3D add-on cards in the beginning, today 3D acceleration is standard on all graphics cards sold including the lowest-end models.
Now imagine nVidia adding the 120-something transistors to their ~700M transistors chips (or alternatively adding a dedicated physics chip on their reference PCBs), making physics processors more widespread. This could mean offloading work of a full CPU core (or second GPU in a SLI-system) to a efficient design dedicated for the task.

The only problem (as with 3D add-on cards back in the days) is that PPUs are not widespread at the moment, so you can't rely on users having one installed. However as back then, games that support dedicated PPUs can go all out in the physics department, same as GPU-supporting games could with astonishing graphics.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

Rarely do pioneers transform the whole landscape by themselves. AMD is wrong that physics on dye with GPUs is dead. Nor is NVIDIA necessarily correct in their timing with this transformation. Much too early to be fully accepted by game developers either, even in light of just one directx API, which initially would provide a minimum return on sales versus additional programming.

I see a lot of 700 pound gorillas beating their chests at the moment, ATI, NVIDIA, and Intel. However, unless the monkey trainer (aka Ballmer) reigns them all in with an API banana, they're just flinging poo at us on the other side of the bars for now.

Re:A physics card is just dual-core for the idiot

[DanJ_UK]

Did you have a turbo button? I miss my turbo button, they just don't make PCs like they used to.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

That is a brick building had individual bricks all stacked on each other. Whenever you hit it with an explosive it would actually crumble like a real building or burn like a real building. That is a lot of calculations which a general CPU isn't the best at.

It's also a lot of wasted calculations for the GPU, because when it's just sitting there, it's still 1000s of objects, and each one needs to be rendered as a separate object. I'm sure that a PPU could be brought to bear on such a task as well, but that means you would essentially require a PPU to process every object before the GPU got to it.

This might make the game really cool for people with a PPU, but it would really suck for those without one.

Re:A physics card is just dual-core for the idiot

[Anonymous Coward]

quad cores arent around the corner, they're here and have been for some time :)
/me strokes his Q6600 lovingly

Re:A physics card is just dual-core for the idiot

[ChaoticLimbs]

Sure, but why do those cores need to be symmetrical? Wouldn't it make sense to have some cores for computationally intense work and others for tasks that require more memory bandwidth and shorter pipelines (and more of them)?

Re:A physics card is just dual-core for the idiot

[gbelteshazzar]

*cuddles his dual core, it's okay baby, you still got a couple of years*

Re:A physics card is just dual-core for the idiot

[sssssss27]

This might make the game really cool for people with a PPU, but it would really suck for those without one. Kind of like how games are really cool for people with GPUs but sucks for those without them.

Re:A physics card is just dual-core for the idiot

[XaXXon]

There's no such thing as an 80487. If you had an SX, you were stuck without a co-processor.

I guess I can stop waiting for Linux support

[Eric Smith]

They've had a Linux version of their SDK for a long time, but it was a software-only version and didn't support their hardware. Given NVidia's lack of enthusiasm for Linux, I suppose if there was any chance that Ageia might have listened to those of us that wanted hardware support on Linux, it's gone now.

Re:I guess I can stop waiting for Linux support

[moosesocks]

Unless something's changed in the past year or two it's been since I stopped using Nvidia, their drivers always tended to be quite good.

They were Binary-only, but they were good in that they were fast, stable, and supported all the major functions of their cards. Hardly half-assed if you ask me.

Havok

[Paul_Hindt]

Considering that I have seen far more games that use Havok than PhysX, I think Intel is at least somewhat in the better position as far as propagation. However, Nvidia could come up with some cool integrated hardware and really push that API to the developers in order to gain some ground. On the other hand consumers would have to bite, and it doesn't seem many have yet caught the physics fever. I have seen Havoc used in numerous console games as well, but AFAIK that's only an API...will Nvidia try to push their own brand of physics hardware along with an API to that sector too? Looks interesting.

Side note: I wouldn't mind seeing more work being done on AI though, possibly leading to AI accelerators in the future.

Re:Havok

[RiotingPacifist]

I have great respect for Nvidia, they really seam to know what theyre doing when it comes to drivers so i can see them getting far if theres a market. Although he same can be said for intel.

AIs in games are rule based AI, so not much specialization can be done, all you could do was produce a simple logic bassed CPU with a high clock, but as AI progressed the AIPU would eventually turn into a slightly simpler CPU (it probably wouldn't need to do floating point calculations but otherwise would be pretty similar).
The time place where i can see an AIPU being useful is game developers got to the point they wanted to introduce sub-symbolic AI (neural net based AI) but there are enough problems with this idea anyway and for the most useful useages of sub-symbolic AI (designing maps*, strategies* or learning** to beat you), a CPU implementation would be fine aslong as it was done between rounds

* they probably wouldnt want to do this as this would mean that game devs could be got rid off and the game AI can make the levels & enemy stratagies and youd get alot more gameplay from a single game as the game would keep making levels
** this is very distant future as even in the lab teaching an AI is very hard and normally hit and miss.

Im trying to think of other game parts that could use a *PU but i other than AI, looks and physics there isnt much to a game.

maybe an anticheat PU that keeps tabs on you without interfering with gameplay, but it does sound abit extreme!

Fab capability...

[Junta]

I would disagree with your characterization of the migration to P4 to core as 'quick'. I would also not declare Intel successfully turning around a product that was competitive across the board with AMD until Core 2, when they pulled in the good instruction per clock and the 64 bit instruction sets all together. It took years for Intel to develop something that *almost* completely dominates the AMD equivalents (one could still make a case for the AMD memory architecture at scale, which Intel will counter with QPI this year). And the clock didn't start ticking until AMD forced their hand.

If it takes a company like Intel years to crank out something like that, a company with debatably the top notch fabrication capabilities in the world, what are nVidia's chances, given that only now they are feasibly able to leverage 65 nm fabrication processes for manufacture of their chips. Fabrication processes aren't everything, but it is a decent indicator of how the cards would be stacked for nVidia going into that market.

I personally would love to see nVidia enter the market with a viable offering, if only because I fear AMD is blowing the situation and the market desperately needs comparable vendors to compete, but I'm not optimistic about nVidia's capabilities.

GPGPU and Nvidia

[volsung]

I'm surprised no one has mentioned CUDA yet, which is Nvidia's existing entry into the world of general purpose GPU computing. So far their target market is mostly dedicated calculations with limited interoperability with OpenGL/DirectX, but I expect we'll see future cards that can partition their compute resources between multiple tasks, like rendering and physics. Hopefully, porting over the PhysX SDK will help grow the GPGPU toolset, and make it easier to use.

(CUDA already transforms the 8800 GTX into quite an impressive array processor. With 128 floating point units and 768 MB of fast, fast memory, this card is chewing up the data-parallel compute tasks I'm throwing at it.)

Re:GPGPU and Nvidia

[Anonymous Coward]

On a side note, CUDA is the plural form of the Polish word for "wonder".

Where we are headed.

[Anonymous Coward]

Where we are headed is a multi-core CPU with dozens of cores; about half general purpose and half specialized for physics and graphics and other purposes. The fabric of the chip will allow the cores to be rewired on the fly into "assembly lines" so that you can rapid fire buffers of data from one core to the next.

There may even be Field Programmable Gate Array or programmable ASIC's inside this same chip that are like computer cores that can be rewired on the fly to be a hardware based protocol encoder or decoder.

Want to be prepared for the future as a programmer? Get a PS3, load Linux on it, and start learning cell programming.

Nvida and AMD were already working on Physics

[BagOBones]

A year ago both Nvidia and ATI/AMD both showed off their GPUs doing HAVOK acceleration equal or better than AGEIA. With ATI claiming to have a 7 month lead... Could this be a catchup move of patent grab by NVIDIA?

http://www.reghardware.co.uk/2006/06/06/ati_gpu_physics_pitch/

Re:Nvida and AMD were already working on Physics

[LingNoi]

How great is the pipeline?

If I have a server with example game running and it's calculating physics, can I then grab back that batch of models I sent off for physics and then send the updated coordinates to the players computers? Or is it just part of the graphics rendering pipeline?

Re:Nvida and AMD were already working on Physics

[Animats]

Can I then grab back that batch of models I sent off for physics and then send the updated coordinates to the players computers? Or is it just part of the graphics rendering pipeline?

It's mostly part of the graphics rendering pipeline. Ageia's "physics engine" is mostly used for particle effects (smoke, fire, rain, snowflakes, etc.) which don't affect the gameplay at all. There were attempts to use it for actual game physics, but the performance was no better than doing that on the main CPU. For particle physics, massive uncoordinated parallelism (the easy case) is possible; the particles don't affect other objects or even each other.

Geographical Ramifications

[BeeBeard]

Does this mean that all boats on the Aegean Sea will now have to use proprietary rudders?

I don't know about you,

[Anonymous Coward]

but I'd prefer the ability to bend the fabric of space and find myself at my destination. Travelling without Moving, as they say.

Talk about a niche market...

[WoTG]

When you can count the number of games that support hardware physics on one hand (actually I made that up, please correct me if I'm wrong), you can be pretty sure that there isn't much volume in the PPU market.

Heck, fewer and fewer PC's come with dedicated GPUs. Integrated video can now handle dual monitor output and HDTV decoding. It's only gamers and graphics designers who need them now.

Re:Talk about a niche market...

[sssssss27]

Heck, fewer and fewer PC's come with dedicated GPUs. Integrated video can now handle dual monitor output and HDTV decoding. It's only gamers and graphics designers who need them now.

Correct me if I'm wrong but doesn't integrated video use an embedded GPU to do the bulk of the work? You can't run Vista with Aero turned on without a GPU so all the Vista PCs shipping with Aero enabled have GPUs in them. I think you meant fewer and fewer PCs come with dedicated graphics cards or high end GPUs. Which is the same reason fewer and fewer PCs require NICs or audio cards. It's getting to the point where embedded solutions are good enough for most people.

Re:Talk about a niche market...

[WoTG]

You're right, dedicated graphics cards is what I meant.
Though, integrating mainstream GPU functionality into the CPU core is only a few years away, IMHO. AMD has said about as much with regards to their "Fusion" core plans. Time will tell.

Re:Talk about a niche market...

[sssssss27]

That's what I figured just wanted to make sure.

Yeah, Fusion should do wonders for the integrated GPU market but we'll always have the dedicated graphics card segment for the enthusiast. The more things we can get on a single chip the better for applications where size is important.

Re:Talk about a niche market...

[NeMon'ess]

Ever since 3D hardware acceleration took off, leaving software rendering behind, the gaming market has been in its own world of accelerator cards.

I think looking back at how 3dfx and glide-only games shows some important similarities. 3dfx managed to capture enough of the gamer market that games were made that would only work on their cards. Maybe only a dozen, but it was still notable that only part of the total market supported having those games. Many of the other games could be run in either Glide or OpenGL. The Glide version usually looked and ran the best.

These days games cost possibly ten times as much to make, and I don't expect a new incarnation of Glide. What I do think is likely is nVidia will add PhysX acceleration to its GeForce 10xxx line of cards. Meaning perhaps half of the video cards bought during that time will have physics acceleration. Around the time the GeForce 11xxx line comes out, games that really take advantage of PhysX will come out. Now hardcore gamers are willing and able to pay $300-600 for a new card. When those games come out and deliver neat effects that only nVidia cards can show off, they'll buy the 11xxx cards or be happy they got 10xxx cards already.

In fact, think back to the original GeForce 256. Do you know what made it different from a Riva TNT 2 Ultra? One had hardware Transform and Lighting (T&L), the other didn't. That's right. When it game to games back then, the last generation card (the Riva) ran games just as well as the GeForce. Back then PC Gamer told people they ought to get the older card if they were on a budget. Then the GeForce 2 came out along with the games supporting hardware T&L. Of course the GeForce line made games look much better than the Riva and at better framerates.

So I think this is a really good move on nVidia's part. They're positioning themselves to devour the high end of the market from ATI when those cards and the games are available. Meanwhile it will probably take ATI at least a generation longer to catch up.

Re:Talk about a niche market...

[VPeric]

When you can count the number of games that support hardware physics on one hand (actually I made that up, please correct me if I'm wrong), you can be pretty sure that there isn't much volume in the PPU market. Here we go, list of games using the PhysX SDK: http://rubux.net/content/view/7/6/. Note that not all of them support the actual card, even if they use the SDK (all of this can be seen on the chart). In any case, while you can't really count 'em on one hand, it's not that much either.

I doubt it

[Anonymous Coward]

The speed of light seems pretty tough to beat.

Re:I doubt it

[Anonymous Coward]

I dunno. Strap a mirror to your back and you'll win the 5k every time.

3D physical desktops

[RiotingPacifist]

I completly disagree, so nvidia dont open source theyre driver, but at the end of the day they release good binaries, I see no advantage to open source drivers for videocards:
*a community isn't going to develop video card drivers as well as the people who make the cards
*a community is much more likely to stall and slow down
*in most cases the fact software is open source doesn't mean much as one company or another has complete control over the product (look at OO)
the only arguments for it are that
*more people will find the bugs (this is a mu point, look at FF, plenty of eyes on code but still plenty of bugs)
*some genius could improve it (look at OO it needs serious work in some areas but nobody bothers)
*there could be spyware in a binary ( stop being paranoid)

the end effect of open sourcing the drivers will be similar to open sourcing secondlife, it just means that its easier to cheat with, no major work has been done on second life but a few people have figured out how to gain unfair advantages ( in the end it will either have to be closed (impossible) or they will waste CPU making sure your not cheating (a real pain) ).

Nvidia are fully committed to linux, they release public betas that are usable for linux, sure they might come a bit later than windows but they do come. Why are ATI open sourcing thier drivers, im guessing because linux users were switching to nvidia as thier drivers worked, either that or they cant be arsed to support linux anymore.

p.s titanic special effects were done on linux-nvidia clusters.

ANYWAY...my point was that this is great news because it means that linux will get fully supported physics cards, meaning some graphics effects can become physical or we can do some 3d physics on the desktop (not sure what we could do maybe throw windows inside the cube? meh i dont even have compiz :'( )
All we need after that are a few opensource to take full advantage, of it.

*Hell even for non linux users this is good news, if nvidia release seperate cards then, linux servers can start taking advantage of server side physics, and allow even physics card less users to benifit!

Re:3D physical desktops

[Anonymous Coward]

Please point me to the FreeBSD/amd64 drivers for my Nvidia card. What do you say, no dice? Oh, but my Intel and ATI hardware works flawlessly because the code is open. Do you understand that there's more in this world than Linux running on x86?

Your comment about Titanic. Man, the guys who did the Titanic FX used SGI workstations for modelling and headless servers for rendering. What made Titanic possible was a lot of talented people and Pixar's PRMan, not a silly Nvidia GPU with a proprietary driver.

Glass

Socks that stay up

[enoz]

Well go and buy some Computer Socks.

Silly AC.

Re:Socks that stay up

[Anonymous Coward]

But Computer Socks fall up. That sounds uncomfortable.

The two architectures are subtly different...

[Joce640k]

Graphics and physics are subtly different tasks. GPUs aren't good at physics and vice versa. A chip which can do both will be a jack-of-all-trades, master of neither. They need to be separate/parallel processors, even if they're on the same chip.

AGEIA's problem is that they're kinda obscure and don't make custom chips.

What NVIDIA brings to the table is a strong brand name and a big manufacturing process. If they can get the price of the PPU down to half of what it is now (by integration into the graphics card and improved process) then they can use their brand name to sell a bucketload.

The trick is to not make it NVIDIA-only because game developers wouldn't buy into that. The trick is to make it run about 25% better on NVIDIA. 25% is enough to swing the buying decision of a hardcore gamer but not enough to scare a game developer off using their SDK.

Re:The two architectures are subtly different...

[TheLink]

The difference between graphics and physics isn't that subtle.

For most games if you turn down the graphics the gameplay isn't supposed to change that much. So people with cheaper video cards can still play the game.

Whereas what happens if you turn down the physics? For the gameplay to not change the crap that's bouncing around can't matter at all.

I'd rather the physics mattered.

But if the physics mattered, people with cheaper physics cards might not be able to play the game.

The game makers won't like that :).

this is good to hear

[sentientbrendan]

physics engines are still relatively simplistic due to the computational difficulty involved. I'd love to see what a good game designer could do with physics capabilities comparable to what modern graphics capabilities look like.

Oxymoronic?

[Anonymous Coward]

>character animation

followed immediately by:

>(ragdoll motion)

I don't think having some algorithm lazily and weightlessly splay out models is at all related to "animation".

On a related note, my CAPTCHA (gotta CAPTCHEM ALL!) for this post was "disgusts".

Are they gonna call it...

[KristoferP]

a GraPhyx card? If they release a combo card that is. No one is gonna buy a seperate physics card, right? But a combo card could enable real time rendering of stuff like this: http://physbam.stanford.edu/~fedkiw/

One step closer...

[Loibisch]

...to realistic simulation of bouncing female anatomy in computer games.

boobies! (. )( .)

The reason AMD wasn't interested

[Alzheimers]

The summary makes it sound like AMD passed on buying AGEIA because they felt that GPU-Based physics acceleration was "dead".

Actually, AMD designed this whole Hyper Transport bus with dedicated hardware acceleration co-processors in mind. In their world, you wouldn't need a dedicated add-in board, just an open HT socket on the motherboard. Then if you want to add the physics acceleration, just pop in the chip. Putting the accelerator on the GPU card would increase the costs of an already expensive board, not to mention throttle the bandwidth over the PCI bus to the video card(s, if SLI) where there's already a major hit because of ever-increasing texture sizes.

What AMD SHOULD be doing is leveraging this technology with their ATI hardware to integrate the graphics into the system, but I guess their OEM partners like selling new cards every 18 months too much to consider going it alone.

AMD should be aiming to take the whole PC platform back to the Amiga days with dedicated co-processors for each multimedia task. HT would allow them to integrate all the hardware seamlessly to the user. Audio, for the most part, has already made the leap onto the chipset (and MSFT did their part by killing off hardware audio acceleration in Vista). But business being what it is, I doubt it'll ever really happen for video. For physics, which is still new and too expensive to be mainstream, it should be less of a risk.

It's not dead,

[Anonymous Coward]

it's just resting. I guess nobody's RTFA for the original ATI vs GPU Physics article.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Just clarifying - PhysX != dead

[ekimminau]

AMD says GPU physics is dead until DirectX 11.

Re:Fuck NVIDIA's proprietary hardware

[graphicsguy]

Okay, so now I can run my hardware-accelerated 3D rendering using an open-source driver on AMD/ATI hardware? Oh, I can't? Why not?!!