Flashpoint 2 And The PPU

[korax 4]

Ah! and how will this work in MP may i ask? Wont this increase data transfer and cause major lag, desinc and latency problems? Or are these abundant physics stuff client side? Even so it would still increase data transfers betwean players and server and back to players considerably, no .

Multiplayer physics will probably be a dumbed-down version of the singleplayer physics to reduce traffic, but with the same end effect.

[Tigershark_BAS 0]

I don't get it...so what will the CPU be doing while the PPu is working.

Isn't this what the CPU is designed for? Isn't it better to simply bet on new CPU technology like dual core processors rather than invest in a piece of technology which may only have limited application, will risk not being accepted and may become outdated with new 64bit dual core processors for home use about 2-3 years away?

[simulacra 0]

The CPU's main job will be AI and coordination etc.

[Antichrist 0]

The CPU's main job will be AI and coordination etc.

I won't be surprised if someone invents new APU (AI Processing Unit) card within couple of years!

[mattxr 9]

im confuesed.. i want a super computer

[simulacra 0]

I won't be surprised if someone invents new APU (AI Processing Unit) card within couple of years!

Well, we ARE heading towards having networked computing in our homes, it's been used in different industrial applications for a long time.

The upside is that you get more processing power since you distribute the workload, the downside is the cost and the need for faster buses...

[Scrub 0]

(Antichrist @ May 24 2005,14:17)

I won't be surprised if someone invents new APU (AI Processing Unit) card within couple of years!

If the workload of the CPU keeps shifting away, IT will be the APU. It's nothing for it (or better yet, the bus itself) to shift skeletal/particle data to the PPU. Lotsa cycles left over for AI to better shoot you with.

[Dwarden 1125]

for me its important how much of AGEIA PPU can be used for other physical engines (Havok , Meqon, Tokamak, ODE etc) ...

i know that it was "discussed" but that was and is still part of rumours ... only time will show truth ...

for Novodex itself i think it will be successfull because Unreal Engine 3 is definitely good runner ...

[gandalf the white 0]

If they use the PPU, then we could open up a .50 on a group of soldiers and have a REAL mess .

that would be like... so... totally... fucked up (Me like! )

[Heatseeker 0]

Ah! and how will this work in MP may i ask? Wont this increase data transfer and cause major lag, desinc and latency problems? Or are these abundant physics stuff client side? Even so it would still increase data transfers betwean players and server and back to players considerably, no .

Multiplayer physics will probably be a dumbed-down version of the singleplayer physics to reduce traffic, but with the same end effect.

Another good reason not to invest 300€ on this i guess, ive heard BF2 features decent physics and ragdolls, for now game2 is far from being done, by the time it comes out CPU will most likely be up for the job of handling decent enough physics, i know this could save developers precious coding time but if it wont even work decently in MP im not interested. Imagine a huge map with alot of players, helicopters, tanks, AT launchers, lotsa stuff being blown up, coliding, really big data transfers, i wouldnt like this to limit the gameplay size, # of players and performance , i see this doing some cool stuff in small games but not really worth it for game2, im not convinced yet...

[gandalf the white 0]

stuff like limbs being seporated, or those VERY small wooden thingy's (read: anything non-lethal) could just be calculated on machines that have the hardware... the game already sends the data of every bullet, etc, so that wouln't be a problem.

now, about the .50 idea....

[MATRA 0]

Quote[/b] ]WE WROTE a couple times before about some wizardry called a PPU (PhysX Processing Unit) and now we've seen the actual hardware in action. Last time we saw this marchitectural miracle we saw the card but all the demos we saw were software rendered.

This time, demos were rendered in real time with all the fancy physics detailed and effects on.

We have to say that seeing hardware in action was breathtaking and we do believe that this is the next step in game evolution.

This hardware add-in card presents 40,000 bones (rigid bodies) on screen at a time, while even Dual core AMD or Intel CPUs can calculate just few hundreds. When it comes to Fluids, PPU can render even more rigid bodies. Your game will look like never before.

AGEIA says sample cards will be ready in Q3. Actually, hardware is ready even now but AEGIA wants to wait for content before launching. The cards should be available in retail sometimes in Q4.

A company spokesman said that in retail those cards should cost between 250 and $300. WE understand Asus will be one of the first companies who will offer those cards in retail.

Its interesting to notice that many of the AGEIA executives comes from 3DFX and those guys sure know how hard it was to convince people to use 3D hardware back in 1995. Now AGEIA has to convince people that games need special hardware to accelerate the Physics in games.

One thing is certain, when you use see this hardware in action you will be convinced. Just a fair warning, it will be very hard to see this hardware in action just by looking at the pictures, you need to watch the video files.

With developers such is Epic, investors such as TSMC and Bank of America, this technology should have a very bright and successful future. µ

http://www.theinquirer.net/?article=23468

This is getting more interesting each day, 3DFX engineers hum... sounds good to me

Too bad the price is still in the same range, but as all the new HW goodies it will drop.

I hope game 2 will have support, at least we could choose if we want to use it or use the traditional physics of the CPU.

Preview: http://www.pcper.com/article.php?aid=140

[shinRaiden 0]

Stretching a CPU's logic to 64bits does not open more lanes on the highway for more entities, it only allows for each lane to be wider. Dual cores do not totally fix that either, as they're essentially two paralell - but independent - roads.

What Ageia can do on their own silicon is do something totally different. They can design a custom ASIC that is specifically tuned to chunk a massive number of similar particle calculations. Appparently they're just in the generic build-to-order ASIC stage, and haven't gone and done anything truely insane like a vast array of vector-matrixed RISC processors etc. If I read the process correctly though, it's operations like this that Cell really shines in, that might be an option down the road.

There's two parts to the system however, the physics SDK and the PPU. If you want to use the PPU you have to use the SDK, but the SDK is one of many options in the SDK market if you do not plan to use the PPU. Haven't seen the commerical terms either, but I assume that it's inline with the other standard offerings out there. You can get a preview copy of the SDK for the cost of a name and email address registration, neither I nor my spam filter have recieved one hit or anything as a result.

Speaking of SDK's though, BIS has been using a licensed SDK from Immersion for some functionality in OFP and VBS, although I'm not sure that they take full advantage of all the libs offered in that SDK. M$'s Dependency Walker doesn't show a whole lot of hooks going into the ifc22.dll.

[RUKH 0]

New PPU clip, the plane crash at the end is soooo nice

http://www.fz.se/filarkiv/?s=&p=mov/physx/

[gandalf the white 0]

the boxes looked a bit buggy, but the rest was too nice to be true...

I really hope OFP2 will have this... *imagines opening .50 on a group of bushes*

[Daniel 0]

It would be great, if not a little unsettling, to actually hesitate before you ripped a group of infantry apart with an M2

[Antichrist 0]

There's two parts to the system however, the physics SDK and the PPU. If you want to use the PPU you have to use the SDK, but the SDK is one of many options in the SDK market if you do not plan to use the PPU. Haven't seen the commerical terms either, but I assume that it's inline with the other standard offerings out there. You can get a preview copy of the SDK for the cost of a name and email address registration, neither I nor my spam filter have recieved one hit or anything as a result.

I believe that would be the thing to kill that PPU eventually. Even if it is extremely successful at launch. Not supporting other physics SDK is an extremely stupid idea. And by the looks of it they're going down that road. I am not really sure what those guys are thinking. They're supposed to be experienced executives who worked on 3dfx. I mean, come on. Don't they realise that if they start supporting ANY physics SDK that wants to be supported and drop price down to 100-150 USD they will make way bigger profit than with their current "business plan".

[gandalf the white 0]

Maybe the PPU guys know something we don't?

If OFP2 supported the PPU technology I'd bet you would scream like a little girl and run for the shop .

The price is a bit high, but like I said: we are entering a new era of gaming! 35 000 objects, compared to the couple hundred objects that most CPU's can currently draw, is a huge leap, and will change the way games are made forever .

Bunker busters, blowing up a dam, once it sounded too far away for any system to handle, now, we're $ 270 away from it!

[shinRaiden 0]

New PPU clip, the plane crash at the end is soooo nice

http://www.fz.se/filarkiv/?s=&p=mov/physx/

The issues seen in the video are due to the game, not the physics system. It's played back in slow motion so you can see the effects, and while the weights are obviously not accurate, that's not the point, the movement is.

[Hidden 0]

I for one welcome the PPU. It is the next logical step in games and simulations, really. As for CPU's, how great is the progress in CPU developement the past couple of years? Not that great. The PC that I own now, I put together some 2 years ago, using a P4 2.8. Today's most efficiant Intel CPU's with a pricetag you and me can afford, is a 3.66. Not much of a difference, really. The focus has been moved onto the GPU for some years now, and it has become insanely powerful and so hot, you can forget about heating expenses during winter (just close the window every now and then) <- joke.

When I first read about the PPU, my initial thoughts were those of a madman who finally realize his fantasy world is real. Or to put it in other words: I had a feeling all along it could come some day.

I welcome the PPU with open arms, and if I can afford it when it is out, I would like to buy one, too. Of course they will be pretty expensive in the beginning; all new technology cost an arm and leg until it goes mass production. Then prices will go down.

Microsoft will be interested in this, I cannot see any valid reason why not. Anything that makes the game and entertainment scene move forward should be of interest to them.

[shinRaiden 0]

There's two parts to the system however, the physics SDK and the PPU. If you want to use the PPU you have to use the SDK, but the SDK is one of many options in the SDK market if you do not plan to use the PPU. Haven't seen the commerical terms either, but I assume that it's inline with the other standard offerings out there. You can get a preview copy of the SDK for the cost of a name and email address registration, neither I nor my spam filter have recieved one hit or anything as a result.

I believe that would be the thing to kill that PPU eventually. Even if it is extremely successful at launch. Not supporting other physics SDK is an extremely stupid idea. And by the looks of it they're going down that road. I am not really sure what those guys are thinking. They're supposed to be experienced executives who worked on 3dfx. I mean, come on. Don't they realise that if they start supporting ANY physics SDK that wants to be supported and drop price down to 100-150 USD they will make way bigger profit than with their current "business plan".

the optimizations for the hardware accelerator is done in the physics libs, not in the end product, although there would need to be an api method to detect and compensate for the varying physics capabilities.

licensing the sdks gives them direct revenue regardless of ppu sales, while ppu sales are limited by market ability to support the cost and the revenue sharing with other entities in the channel. supporting other sdks on the ppu would theoretically increase low-margin chip sales, but negatively impact high-margin sdk licenses.

conversely, vendors such as microsoft (directx), nvidia (cg, nv_gl_extensions), and others provide their sdks free of charge to promote usage of their core product. ageia's in a bit of a sticky spot here, novodex is nowhere near being the incumbent physics sdk like others mentioned here. but you have to use novodex to use the ppu. to change a system in development is a non-trivial matter. ofp1/vbs1 use a partial package from immersion "we sued sony" fame, and unless jordan man derek is hiding something mind boggling juicey about the physics system, it's rather difficult to explain to the accounting dept why you would want to abandon an expensive licensed lib, splurge on another, and force a major calendar slippage, unless jmd's already been using novodex for some time...

[Tigershark_BAS 0]

Found this today. John Carmack's take on the PPU at Quakecon.

Physics acceleration's prospects

Carmack said he considers the prospects for the upcoming physics acceleration chip on the PC iffy, because physics presents a very fundamental problem that graphics doesn't have: it isn't easily scalable for level of detail. Either an object in the game is a true physics object with which other objects can interact, or it isn't. Carmack predicted this constraint would lead to a number of physics-accelerated titles where acceleration affects only elements, such as flowing water, that are peripheral to the core gameplay experience.

Another of his concerns about physics acceleration was speed, not in the sense of peak processing throughput but in terms of the immediacy of real-time interactions. Carmack recalled that the first pre-3dfx graphics chips made Quake feel slower due to lag between user input and visual output. He worried that the first generation of physics chips could cause similar problems, leaving us with games that are fragile and slow. He also happily conceded that he is more of a graphics guy than a physics guy, and admitted that his worries about the bar being raised in physics are probably similar to the worries other have had about his own standard-setting work in graphics.

[chipper 0]

Flashpoint 2 and its websote was removed from flashpoint1985.com! They are maybe changing it to look kickass and more info

[Ryujin 0]

Check out these fluids! HOLY @$#%!

http://dlx.gamespot.com/pc/eyeofthestorm/moreinfo_6125715.html

[Pyronick 21]

for me its important how much of AGEIA PPU can be used for other physical engines (Havok , Meqon, Tokamak, ODE etc) ...

i know that it was "discussed" but that was and is still part of rumours ... only time will show truth ...

for Novodex itself i think it will be successfull because Unreal Engine 3 is definitely good runner ...

Yeah but I hope that Havok dies!!!!!

ODE and Meqon sound great though, but a competitor like NovodeX which also is powered by the Open Dynamics Framework (OpenDE is too) and also features real-time physics calculation (like Meqon does)...

Next to that its fluid simulation looks promising, and metal bending too...

Full-Scene Dynamic Destruction is the best though :D

VBS1 has Dynamic Destruction too but not as advanced as this so I'd say... Bring on the PhysX PPU and NovodeX (and of course the Game2 physics with full PPU support)...

Quote[/b] ]The upside is that you get more processing power since you distribute the workload, the downside is the cost and the need for faster buses...

Not necessary, the current PCI bus and PCI-X bus can handle it perfectly.

Maybe the good 'ol ISA bus would be quick enough...

Transfer wouldn't be that huge, it is only about the coordinates of the bones of an object.

Nothing more... Maybe we could drop Pixelshader on water too, since the PPU does the terrain leveling of the water surface :D