A Closer Look At PhysX: Our Take On The PPU
by Derek Wilson on March 11, 2005 12:08 PM EST- Posted in
- CPUs
Game Physics and the PhysX PPU
One of the properties of graphics that made the feature a good fit for a specialized processor inside a PC is the fact that the task is infinitely parallelizable. Hundreds of thousands, and even millions of pixels need to be processed every frame. The more detailed a rendering needs to be, the more parallel the task becomes. The same is true with physics. As with the visual world, the physical world is continuous rather than discrete. The more processing power we have, the more things we can simulate at once, and the more realistically we can approximate the real world.
But as we’ll see, there is more to it than simply parallel compute power.
In the beginning there was collision detection, and it was good. And then came some semblance of gravity. Over the years, more and more objects in the game world have become interactive and affected by the world around it. Now the game physics world has culminated in the ability to bowl for oil drums with floppy dead people.
Fast, efficient, rigid body simulation has been well developed over the past couple years. A rigid body is basically a single unit that does not deform due to physical interaction. Think of two billiard balls slamming into each other: the balls retain their shape. If one were to drop a billiard ball on a beach ball, we would see a much different reaction: the beach ball would contract and expand a great deal. In current games, we would treat the beach ball, and indeed everything else, as a rigid body object.
Any rigid body object can have a number of forces acting on it at a time depending on the complexity of the simulation. Each object also has its own properties that are used to determine how these forces affect the object. This can scale from simple (mass and shape) to complex (like data to describe how shape and mass distribution affect angular velocity) depending on how accurate the developer wants to make things.
Not much interesting happens if a rigid body is sitting on flat ground and not moving with nothing touching it. Interesting things start to happen when either the user interacts with an object, or multiple rigid bodies interact with each other. We can easily see how these interactions could get complex. Think of bowling or billiards. Now think a little more complex. What happens when a rocket hits a brick wall and all the bricks go flying? Currently, games generally use tens to about a hundred rigid body objects in any given scene. This is somewhat limiting when the game calls for large scale effects or destruction.
So, here’s the real question. Why won’t Intel’s vision of multi and many-core processors be good enough to handle what a discrete PPU could handle?
Of course more parallel processing power will help no matter how it comes. But the deeper issue is data movement. The example AGEIA gave us to think about was a huge pile of bricks; when you push down on one brick near the top, forces are transferred to all other bricks in the stack. This may not be difficult on the scale to hundreds, but how about 30000 objects in a stack? How about keeping track of that while handling deformable (soft body) objects, fluids, and all collision detection in the scene?
The was all this needs to be handled is not simply with lots of parallel independent floating point power, but with lots of parallel floating point power connected by huge bandwidth. The fact that some initial Intel dual core chips will have to go off chip and back on to communicate, not as much performance is gained as possible. Certainly more parallelism is better no matter what, but it’s the high bandwidth that clenches the deal.
AGEIA’s background is in switched fabrics and networking hardware. The one of the keys to their chip is in the way PhysX is able to move high volumes of computed data internally. In order to find the forces on a brick in the middle of a stack that’s been kicked, the forces on all the bricks around it need to be updated.
Getting data into and out of the chip as fast as possible is important as well. For now, AGEIA is sticking with GDDR3 in order to benefit from the high volume (and lower costs) generated by the graphics market. It would be possible for the PhysX to benefit from higher bandwidth solutions like XDR, but for now it is important for AGEIA to minimize the cost/benefit ratio in order to succeed. The company philosophy at the outset is to follow what the graphics market does in terms of on card RAM.
We really don’t know a great deal about the intimate details of the architecture, but a light weight parallel floating point with lots of communications is a good start. We’ve had several guesses at how the hardware works that have been confirmed wrong. But to paraphrase Edison, eliminating all incorrect paths leads to the goal.
It does seem obvious that all physical properties of objects can be uploaded to the hardware initially (like a graphics card does with textures and such) and manipulated/updated every time something changes. We’ll be bringing out as many details as we can as soon as we are able.
Facebook
Twitter
RSS
70 Comments
View All Comments
REMF - Saturday, March 12, 2005 - link
if ageiaieiai are sensible they will license this tech to ATI/nVidia as quickly as they can.they will have about 12 months lead time after which if the big GPU makers consider physics acceleration a 'go'er' then they will integrate it into their GPU's at mafginal cost.
.13u = 225m transistors
.09u = 350m+ transistors
.065u = 550m+ transistors
there is easily room to squeeze 120m trannies of PPU onto a GPU die created in 2006.
remember that 3d started off as a separate product, but economies of scale and efficiency of operation eventually demanded that 2d and 3d become one. i don't see any reason why physics acceleration should be any different.
hoppa - Saturday, March 12, 2005 - link
Wow this is the most exciting thing I've seen in a really, really long time."But like plasma television (which has been around for decades), just because good technology exists doesn’t mean vendors and consumers will adopt it."
Whoa, wait, how many thousands of dollars do plasma TVs cost? That is a very different thing! Especially since there is a very suitable replacement for about 1/10th the price (a good ol' CRT).
RockHydra11 - Saturday, March 12, 2005 - link
Just one more thing to make platforms more expensive, create fanboys of band X and Y and shove hype and BS down our throats.Poser - Friday, March 11, 2005 - link
Sounds great for a console, where the developers can expect 100% of systems to have it. The PC? Eh...I've got to wonder how well something like this could be pitched to animation studios. Are the physics calculations taking up any significant manpower/processor time, or does the image rendering dwarf everything else?
One hypothetical add-in card I'd love to see instead of physics is hardware dedicated to speech recognition. Not a clue what sort of hardware would be required to improve the current state of the or even if dedicated hardware would be *able* to help.
archcommus - Friday, March 11, 2005 - link
The idea is great, it has the same advantage as the GPU has for graphics, but honestly, who the hell is going to want to have to have a PHYSICS card? A sound card is fine, as you can use it for a number of things. A graphics card is fine, because it powers the graphics for everything, not just games. But a physics card would be something that you must add to your computer JUST for gaming, and it would not benefit anything else at all.Great theory, but sorry, I don't want to plunk down ANOTHER $200 for every system I build just so I can play the latest games decently.
Kensei - Friday, March 11, 2005 - link
On an overall editorial note, could the AnandTech staff please define key acronyms (particularly ones in the title of the article) at the beginning of each article. What the hell is a PPU?Gunbuster - Friday, March 11, 2005 - link
Devs. wont even code dual screen into games, and you expect them to support a esoteric board 0.1% of users will have?The only chance is if they strike a deal with one of the next gen consoles
AnnihilatorX - Friday, March 11, 2005 - link
Re REMF #28 >Well new products are always expensive even they can be cheaply made. There's just no competition when you offer a brand new product. I doublt it though would be too expensive. Let's see.
<
hmm modelling clothings that can be torn apart... <- been thinking this for 5 min now. Hmm I need to clean my mind for a bit.
SDA - Friday, March 11, 2005 - link
regarding plasma: Plasma screens are inherently hideously expensive. That's actually a design issue. (Okay, so it's an issue with LCDs too, but refined manufacturing techniques and sufficiently wide-scale manufacturing can mostly solve that problem.) The fact that plasma screens remain expensive to manufacture after being around for so long is evidence of this-- most old tech still in use is relatively cheap simply because of all the manufacturing and design experience we have.A better comparison, IMHO, would be SCSI. Good technology, lots of potential, but it was never adopted by a wide consumer base (although it's always been popular in systems that absolutely need disk performance, for obvious reasons), so prices on SCSI controllers and drives never dropped enough, so it wasn't adopted widely, so prices didn't drop, and so on.
But enough nitpicking. I do think PPUs have incredible potential... honestly, I believe that accurate physics simulation is the single biggest obstacle on the Path Towards Game Realism(tm) right now. It's pretty easy for the brain to work around the fact that a character model isn't a real live person (especially if you can't tell the difference when you squint), but objects not acting in a manner consistent with physical reality, well...
Denial - Friday, March 11, 2005 - link
I bet Virtual Valerie 12 will FVCKIN ROCK with this new physics chip! Just imaging poking her so damn hard you can see it pokin at her belly.I can see the ads now.
"Virtual chicks on trampolines never looked so real."