ATI Physics

[Fistandantilis]

I thought I read somewhere that the 1900 series cards had some sort of physics engine, am I way off on this.

[quadnad]

There's a physics processing forum.

Check there for more details.

[Lazy_Moron]

No, its like the SLI Pysics. You need Crossfire to use it. I think its a waste of GPU power. This is why we are having dedicated physics cards. GPU's are meant to draw are sexy looking graphics not calculate are physics...

[Fistandantilis]

There's a physics processing forum.

Check there for more details.

oh ya, sorry I forgot. man and I finally had a physics question too.

[Brent_Justice]

No, its like the SLI Pysics. You need Crossfire to use it. I think its a waste of GPU power. This is why we are having dedicated physics cards. GPU's are meant to draw are sexy looking graphics not calculate are physics...

That is not true, you don't need CrossFire to use it.

[Fistandantilis]

OK well I am going to the physx section. I should be able to find out all the details there.

[BBA]

That is not true, you don't need CrossFire to use it.

...And, if ATi accellerates the same functions as Ageia, the ATi card is faster then Ageia by a factor of three or more times.

[Cadaver]

...And, if ATi accellerates the same functions as Ageia, the ATi card is faster then Ageia by a factor of three or more times.

I don't follow your reasoning. Could you elaborate please?

[Trimlock]

No, its like the SLI Pysics. You need Crossfire to use it. I think its a waste of GPU power. This is why we are having dedicated physics cards. GPU's are meant to draw are sexy looking graphics not calculate are physics...

incorrect, you don't need crossfire enabled for it to work, you can have a secondary GPU do physic's computations completely and have no impact on video performance

the secondary GPU can be anything, i'm not sure how far back it goes to supported units, but i think it might scale all the way back to the R420 cores, either way the set up ATi has built would work with an X1900XTX and a X1300pro, have the X1300 work as the aiding physic's GPU and have the X1900 be your main GPU

[ElMoIsEviL]

Exactly...

But you need an x1600 or higher. the x1300 doesn't support the functions required.

I have any of these I could use x1800XL, x1800XT & x1900XT to pair up with my x1900XTX.

But I am most probably going to ALSO buy an Ageia card as soon as they're available. I received 3K from my taxes I overpaid... so yeah.

No Crossfire though. I find it useless as I game on a 19" LCD @ 1280x1024.

[BBA]

I don't follow your reasoning. Could you elaborate please?


Sure, from here: http://www.megagames.com/news/html/hardware/atiageiaandnvidia-itsphysics.shtml


in-game physics relies heavily on floating point arithmetic, its R580 architecture is ideally suited to it since it features 48 pixel shaders and suggests that it has 375 GFlops per card available for such calculations. This number compares favorably to the 10 GFlops available in the fastest widely available CPUs and the 100 Ageia will offer. Other aspects of ATI architecture such as dedicated branching logic, unified shader units and a 3:1 shader/pipeline ratio also offer advantages when performing physics calculations.

[Anarchist4000]

I'm thinking this is going to be much more useful when everyone goes to upgrade their video cards down the road. You can get a brand new video card and then stick the old 1800/1900 or whatever in possibly a 4x slot and use it like a physics processor. It would still be significantly faster than the CPU at processing a lot of physics operations and give you something to do with the card.

[Fistandantilis]

OK so I really want to get a PPU but I am not going to go there till vista is out and things are rolling smoothly.
My main concern is getting good frames with GRAW with all the bells and whistles... at 1680x1050, thats all I hope I aint asking too much.

Man I really cant wait for this game, ghost recon is the game that started it all for me, after the NES I stopped video games till I saw recon, now I am a game whore! and recon comming back is making me itchy.

[Trimlock]

But you need an x1600 or higher. the x1300 doesn't support the functions required.

thanks for clearing that up

[Cadaver]

Sure, from here: http://www.megagames.com/news/html/hardware/atiageiaandnvidia-itsphysics.shtml

Ah, I see. However, that wouldn't necessarily mean an X1900 would be 3 times faster than an Ageia PPU, as the GPU would have to expend most of its efforts generating graphic content. Still, I'm very interested to see the direction that Ati will take. :D

[PWMK2]

Ah, I see. However, that wouldn't necessarily mean an X1900 would be 3 times faster than an Ageia PPU, as the GPU would have to expend most of its efforts generating graphic content. Still, I'm very interested to see the direction that Ati will take. :D

Yup. This is why you don't have CPUs handle game graphics any more, despite the fact that your average CPU is lightyears more powerful than your average GPU. :D

[BBA]

Ah, I see. However, that wouldn't necessarily mean an X1900 would be 3 times faster than an Ageia PPU, as the GPU would have to expend most of its efforts generating graphic content. Still, I'm very interested to see the direction that Ati will take. :D


ATi is not driving just balancing Physics and Graphics on the same card, but they are driving letting you dedicate a graphics card to physics, such as when the X1900XTX is old, and you have the X2900XTX and such as your primary DX10 card with Vista...

At that point yes, it will produce 3 time the physics calculations as the Ageia PPU.

[Terra]

ATi is not driving just balancing Physics and Graphics on the same card, but they are driving letting you dedicate a graphics card to physics, such as when the X1900XTX is old, and you have the X2900XTX and such as your primary DX10 card with Vista...

At that point yes, it will produce 3 time the physics calculations as the Ageia PPU.

You are still talking wanna-be physics...
HavockFX has NOTHING to do with REAL gameplay Physics...

Unless you can show my how the pixelpipes can relay physcis information back to the CPU, wich is REQUIRED in order to have GAMEPLAY physics, your not making any sense...
Pixelpipes are for post-processing, not feedback...
You tried the same BS in this Thread:
PPU - Forum: Let me explain the business model here... (http://www.hardforum.com/showthread.php?t=1037928&page=1&pp=20)
But you seem to have left you own thread, to come here instead, and continue the same path of false information

Why don't you come back and make that claim there?
You are more than welcme, and I promise you that you will get qualified feedback...
Instead of playing a ATI-PR machine...

This thread should be moved to the PPU/Physics forum

Terra - Stop spreding your FUD, please...

[Digital Viper-X-]

ATi is not driving just balancing Physics and Graphics on the same card, but they are driving letting you dedicate a graphics card to physics, such as when the X1900XTX is old, and you have the X2900XTX and such as your primary DX10 card with Vista...

At that point yes, it will produce 3 time the physics calculations as the Ageia PPU.

where do you get that 3X number?

[Terra]

where do you get that 3X number?

He is taking pure GFlops numbers and comparing them cross technology/platform :rolleyes:

Terra...

[BBA]

where do you get that 3X number?

Did you not read the article I linked? Physics is based on raw floating point calculation, and since ATi is currently over three times faster than Ageia, it implies better than 3X physics capability. Wether that pans out in the end...well, it's a waiting game.

ATi = 375 GFLOPs
Ageia = 100 GFLOPs

The question remains, does ATi want to support Ageia's physics SDK or make their own...and that depends on wether Ageia wants to let ATi either license it -or- if Ageia wants ATi to crush them in the market place with better hardware AND better physics (May not be better, but it would be a standard as soon as ATi includes it in the drivers).

[Fistandantilis]

fellas, fellas, please dont argue, I by no means wanted to start bickering with this thread... man no I dont even want a PPU.

[Terra]

Did you not read the article I linked? Physics is based on raw floating point calculation,...
That is one of the largest oversimplifications, I have ever heard...

and since ATi is currently over three times faster than Ageia, it implies better than 3X physics capability. Wether that pans out in the end...well, it's a waiting game.

Again, one of the worst oversimplifications I have heard...

ATi = 375 GFLOPs
Ageia = 100 GFLOPs

Again comparing pure GFlops numbesr cross platfom/technology is stupid.
Hell even when comparing CPU's it stupid.
Microsoft's Xbox 360 & Sony's PlayStation 3 - Examples of Poor CPU (http://groups.google.com/group/alt.games.video.sony-playstation2/msg/62ff83d96ea78ea9?hl=en)

Secondly, both floating point power numbers refer to the whole system, CPU
and GPU. Obviously a GPU's floating point processing power doesn't mean
anything if you're trying to run general purpose code on it and vice versa.
As we've seen from the graphics market, characterizing GPU performance in
terms of generic floating point operations per second is far from the full
performance story.

And:
Another way to look at this comparison of flops is to look at integer add
latencies on the Pentium 4 vs. the Athlon 64. The Pentium 4 has two double
pumped ALUs, each capable of performing two add operations per clock, that's
a total of 4 add operations per clock; so we could say that a 3.8GHz Pentium
4 can perform 15.2 billion operations per second. The Athlon 64 has three
ALUs each capable of executing an add every clock; so a 2.8GHz Athlon 64
can perform 8.4 billion operations per second. By this silly console
marketing logic, the Pentium 4 would be almost twice as fast as the Athlon
64, and a multi-core Pentium 4 would be faster than a multi-core Athlon 64.
Any AnandTech reader should know that's hardly the case. No code is
composed entirely of add instructions, and even if it were, eventually the
Pentium 4 and Athlon 64 will have to go out to main memory for data, and
when they do, the Athlon 64 has a much lower latency access to memory than
the P4. In the end, despite what these horribly concocted numbers may lead
you to believe, they say absolutely nothing about performance. The exact
same situation exists with the CPUs of the next-generation consoles; don't
fall for it.

I could turn this around, and use the same "logic" as you:
ATI = 49.6 GB/sec bandwith
AGEIA = 2 TB/sec bandwith

AGIEA = 40x faster than ATI...

But that would be stupid..since it cross platform/technology...

The question remains, does ATi want to support Ageia's physics SDK or make their own...and that depends on wether Ageia wants to let ATi license from them or it Ageia wants ATi to crush them in the market place with better hardware and better physics.

Again, I would loooooove to hear how pixelpipelines can feed collision data back to the CPU? :)

Terra...

[Terra]

fellas, fellas, please dont argue, I by no means wanted to start bickering with this thread... man no I dont even want a PPU.

I am not arguing, I am using arguments.
But someone in this thread has been feedning a lot of misinformation.
And that not good for the thread...or you..or any of us ;)

Terra...

[Trimlock]

Terra i agree with you, but ATi is not single handidly only supporting HavockFX

You are still talking wanna-be physics...
HavockFX has NOTHING to do with REAL gameplay Physics...

might not be as fast as the SLI implementation when the havock engine is used but it should work across all platforms

i'm just waiting for it to come out :D

[Terra]

Terra i agree with you, but ATi is not single handidly only supporting HavockFX



might not be as fast as the SLI implementation when the havock engine is used but it should work across all platforms

i'm just waiting for it to come out :D

I still don't understand how they are going to violate the laws of physics (pun intended) and make the pixel pipelines feed back collision data to the CPU? ;)

Terra...

[quadnad]

I'm still waiting for something other than a demo for a game thats arriving in late 2k7 to tell me that this is all worth it...

[Terra]

I'm still waiting for something other than a demo for a game thats arriving in late 2k7 to tell me that this is all worth it...

It's worth it :p

Terra - Done :D

[quadnad]

It's worth it :p

Terra - Done :D

lol, newegg, HERE I COME!

[Terra]

lol, newegg, HERE I COME!

Sorry, couldn't resist ;)
But just think how geared up people where about the low physics in HL2...
Multiply that by a VERY big factor...and that is what PPU's will bring to the table ;)

Terra...

[Trimlock]

I still don't understand how they are going to violate the laws of physics (pun intended) and make the pixel pipelines feed back collision data to the CPU?

hehe, either way i'm sure they can handle it through some emulation, which would bring performance down, but this isn't replacing a PPU, its pretty much just creating one in theory, if this does do the job with less speed through which ever process they come up with i'm still a happy camper

[Terra]

hehe, either way i'm sure they can handle it through some emulation, which would bring performance down, but this isn't replacing a PPU, its pretty much just creating one in theory, if this does do the job with less speed through which ever process they come up with i'm still a happy camper

Emulation?
That won't cut it.
For gameplay physics to operate, collision data have to be feedback to the CPU.
The problem is that once data enters the pixel pipelines, it's out of reach of the CPU.
You can't "emulate" a new GPU-stucture that is bi-directional...
The pipelines are a one way street...

Terra...

[PWMK2]

I think I've figured it out: get rid of the CPU. :D

[Trimlock]

Emulation?
That won't cut it.
For gameplay physics to operate, collision data have to be feedback to the CPU.
The problem is that once data enters the pixel pipelines, it's out of reach of the CPU.
You can't "emulate" a new GPU-stucture that is bi-directional...
The pipelines are a one way street...

i'm not trying to pretend to know what ATi or Nvidia has in store, but they wouldn't exactly put in development time into it if it didn't produce ;)

[Majin]

That is one of the largest oversimplifications, I have ever heard...

Again, I would loooooove to hear how pixelpipelines can feed collision data back to the CPU? :)

Terra...

OWNED!
Nice counter points Terra.

But I have to take this a different direction
you're SOOOO WRONG
P4 > A64
Xenos > Cell
X1800 > X1900XTX
nVidia AF > ATI AF
Green > Blue
Horse > Dog
69 > 68
At > The

I think you know your roll!


Majin - Shaking my head at Terra for 6+ Months!

[rincewind]

Forgive me if I'm being an idiot, but doesn't the current PhysX have to feed data back to the CPU over the PCI bus? 2Tb/s internal bandwidth is all well and good, but the PCI bus only offers ~132Mb/s (some of which is already eaten up). PCI-E x16 gives a more respectable 6.4Gb/s, and the bus itself is capable of double that (PCI-E mobos normally have 32 lanes afaik) Am I missing something, or does PhysX have a major achilles heel right there? The data can be moved internally at lightspeed, but the most it can afford to transmit is a meagre 4 megs per frame (to maintain 30FPS).

[BBA]

I could turn this around, and use the same "logic" as you:
ATI = 49.6 GB/sec bandwith
AGEIA = 2 TB/sec bandwith

AGIEA = 40x faster than ATI...

But that would be stupid..since it cross platform/technology...



Again, I would loooooove to hear how pixelpipelines can feed collision data back to the CPU? :)

Terra...


Think about what you said. Then explain how a PCI connected card is going to have a higher bandwidth back to the CPU than a PCI-E connected card is. The video card has a great advantage here as well, now that you bring it up. :D

The 'internal to the GPU/PPU' memory bandwidth means nothing, it's the processing that counts as well as getting the data back to the CPU and video card.

Thats why a reprogrammed X1800 is able to process video compression/rendering in seconds compared to the P4 doing it in almost an hour. (You can find that article link here if you look). All it takes is reprogramming the shaders and ATi is already making the drivers to do it.

[BBA]

Forgive me if I'm being an idiot, but doesn't the current PhysX have to feed data back to the CPU over the PCI bus? 2Tb/s internal bandwidth is all well and good, but the PCI bus only offers ~132Mb/s (some of which is already eaten up). PCI-E x16 gives a more respectable 6.4Gb/s, and the bus itself is capable of double that (PCI-E mobos normally have 32 lanes afaik) Am I missing something, or does PhysX have a major achilles heel right there? The data can be moved internally at lightspeed, but the most it can afford to transmit is a meagre 4 megs per frame (to maintain 30FPS).

Exactly my point.

..looks like Terra has been owned once again.

[BBA]

Emulation?
That won't cut it.
For gameplay physics to operate, collision data have to be feedback to the CPU.
The problem is that once data enters the pixel pipelines, it's out of reach of the CPU.
You can't "emulate" a new GPU-stucture that is bi-directional...
The pipelines are a one way street...

Terra...

The thing you are not recognizing is the GPU does not emulate...it reprograms teh alrogithyms and does it in hardware.

By your logic, a GPU emulates video rendering as well...

Chew on that a while.

[BBA]

And...where is anyone mentioning Havok even relevant when we are discussing real physics?

But, if you were curious, both ATi and nvidia are working havok support in, as an added selling point. (Havok is nothing but sequenced events that look like physics but are not real interaction calculations like true physics calculations are).

[Cypher19]

Forgive me if I'm being an idiot, but doesn't the current PhysX have to feed data back to the CPU over the PCI bus? 2Tb/s internal bandwidth is all well and good, but the PCI bus only offers ~132Mb/s (some of which is already eaten up). PCI-E x16 gives a more respectable 6.4Gb/s, and the bus itself is capable of double that (PCI-E mobos normally have 32 lanes afaik) Am I missing something, or does PhysX have a major achilles heel right there? The data can be moved internally at lightspeed, but the most it can afford to transmit is a meagre 4 megs per frame (to maintain 30FPS).

While I'm not familiar with the PhysX API, you'd only need to send back a few floating point numbers per object in most cases: 3 for position, 3 for orientation. For boxes tumbling about, you'd have to have about 175,000 boxes to saturate the PCI bandwidth at that rate. There simply isn't a lot of data that needs to be sent back to the CPU in most cases. Of course, PCIe would be less limiting, no?

Emulation?
That won't cut it.
For gameplay physics to operate, collision data have to be feedback to the CPU.
The problem is that once data enters the pixel pipelines, it's out of reach of the CPU.
You can't "emulate" a new GPU-stucture that is bi-directional...
The pipelines are a one way street...

No they're not. The pipelines generate data which is sent to a location in the GPU's memory. From there, it's not that hard to copy data from the GPU, over the AGP bus, and into system memory to get the results you need. Yes, it stalls a bit when you try to do something like that on D3D, but ATi is developing their physics API such that they're able to talk and interact VERY directly with the GPU, and so the stalling that would normally occur can be easily avoided.

[Terra]

i'm not trying to pretend to know what ATi or Nvidia has in store, but they wouldn't exactly put in development time into it if it didn't produce ;)

Ever heard of markeeting bullshit before? ;) *L*

Terra...

[Terra]

OWNED!
Nice counter points Terra.

But I have to take this a different direction
you're SOOOO WRONG
P4 > A64
Xenos > Cell
X1800 > X1900XTX
nVidia AF > ATI AF
Green > Blue
Horse > Dog
69 > 68
At > The

I think you know your roll!


Majin - Shaking my head at Terra for 6+ Months!

*LOL*

Terra - Nice humor ;) :D

[Terra]

Think about what you said. Then explain how a PCI connected card is going to have a higher bandwidth back to the CPU than a PCI-E connected card is. The video card has a great advantage here as well, now that you bring it up. :D

No it dosn't:
One of the biggest characteristics of the PPU is its memory architecture. It has 128bit memory interface for external memory, but has no internal cache memory.

“We don’t have cache memory hierarchy of any kind. This is very important because traditional cache is not suitable for physics," says Hedge.

PPU has no structure such as CPU cache that is synchronized with external memory by the set-associative method and updates automatically. It’s because in physics simulation it has little data locality. They say memory cache hierarchy is more trouble than it’s worth.

“In CPU and GPU, data has locality. But in physics not, as it has to do random access to many objects. Data structures are totally different” says Nadeem Mohammad, who moved from a GPU vendor to AGEIA.

Still PPU has large internal memory in itself. It has various internal memories instead of cache, and has the organization that does explicit and programmable transfer between internal and external memories.

The patent explains memories such as dual-bank Inter-Engine Memory (IEM) connected to VPU, multi-purpose Scratch Pad Memory (SPM), DME Instruction Memory (DIM) which does instruction queuing, and so on. Hedge suggested that those memories in the patent are in the actual implementation by saying “they are probably included” in PPU.

Among those memories IEM is used in the way that looks like traditional data cache. According to the patent, DME loads a data set required for operation of processing units into IEM explicitly. Unlike cache memory, low-latency access is possible in IEM and apparently it could implement a large number of I/O ports. As the result, it could achieve huge internal memory bandwidth.

“One of the important factors in a physics architecture is it requires huge on-chip memory bandwidth. Our PPU has 2Tb(Tera-bit)/sec on-chip memory bandwidth," says Hedge.

In short, removing complicated cache control made it possible that PPU has L2-cache size internal memory with L1-cache latency and huge bandwidth, and it’s suitable for physics algorithm according to them.


BTW...i found the "source" of your claims:
http://www.firingsquad.com/hardware/ati_physics/page2.asp

As for performance of the PPU vs. the GPU, the issue is one of efficiency and processing power. We think that even if Ageia’s PPU has 100% processing efficiency and ATI’s GPU has 80% efficiency, if the PPU only has 100 Gflops of processing power, and we have 375 in our GPU, we’ll still have a higher performing solution (those numbers are just examples, not proven stats, but I think you can understand how we’re looking at this).

Your numbers aren't even facts...come on :rolleyes:

The 'internal to the GPU/PPU' memory bandwidth means nothing,

That's a LIE, read my first quote!

it's the processing that counts as well as getting the data back to the CPU and video card.

Thats why a reprogrammed X1800 is able to process video compression/rendering in seconds compared to the P4 doing it in almost an hour. (You can find that article link here if you look). All it takes is reprogramming the shaders and ATi is already making the drivers to do it.

Comparing a CPU to a GPU is just a stupid as comparing a GPU to PPU...
Let's recap:

Your numbers are NOT facts...
GFlops mean nothing in terms of acutal prefomance.
The PPU is buildt with an internal bandwith of 2 TB/sex..to keep track of those darn physics objects, that don't follow a fixed path...

Terra - So far...you GPU-PR isn't doing good..

[Terra]

The thing you are not recognizing is the GPU does not emulate...it reprograms teh alrogithyms and does it in hardware.

By your logic, a GPU emulates video rendering as well...

Chew on that a while.

No, they are adding more POST-PROCESSING effects in the GPU...

Terra - Chew on that on...

[Anarchist4000]

Moving data off the video card is a relatively simple process. Right now it's as simple as taking a screenshot. If the drivers were updated to expect the data to frequently being sent back to system memory any delays could likely be avoided. Information could then be sent just as fast as it is received.

As far as comparing GFLOPS the numbers could be kind of close assuming the ATI number is only counting pixel processing power. With the 3 ALUs on the 1900's they could probably tear through a fair amount of data.

The key thing with physics is doing everything in parallel which is how video cards prefer to do things already. Collision is where things can get tricky because a logic comes into play and you start comparing to a bunch of stuff. Implementing a true more realistic system could be kind of rough.

Using a R2VB style approach most basic acceleration, velocity, position based calculations could be done on the GPU. Then the required data dumped off for the CPU to process. Then for the next frame the CPU would feed back data based on what collisions occured. Doing comparisions on a CPU would be much easier than running FP based math on it non stop.

Keep in mind for most games you don't need 100% accurate results. You just need close approximations so even a rough result would be beneficial. The key to good collision has always been intelligently narrowing down what needs tested against.

[BBA]

No it dosn't:


BTW...i found the "source" of your claims:



So...you finally decided to read what I posted above.



Comparing a CPU to a GPU is just a stupid as comparing a GPU to PPU...
Let's recap:

Your numbers are NOT facts...
GFlops mean nothing in terms of acutal prefomance.
The PPU is buildt with an internal bandwith of 2 TB/sex..to keep track of those darn physics objects, that don't follow a fixed path...


And you tell me just what is the GPU internal bandwidth? All you seem to quote is the video cards memory bandwidth, and considering the PPU has low speed memory, I just don't think your comparing apples to apples here. The internal PPU bandwith could be lower than the GPU internal bandwidth...but since those numbers havent been posted, I still say your going on BS marketing crap.

[Terra]

So...you finally decided to read what I posted above.

I didn't need to read anything I ahdn'r read before to know you where waaay of.
But let me hear again:
Why are you using numbers that are not facts as "arguments"?

And you tell me just what is the GPU internal bandwidth? All you seem to quote is the video cards memory bandwidth, and considering the PPU has low speed memory, I just don't think your comparing apples to apples here. The internal PPU bandwith could be lower than the GPU internal bandwidth...

And then reality sets in:
http://www.pcper.com/article.php?aid=225&type=expert

"So how does AGEIA’s first generation PPU (physics processing unit) address these requirements and keep the future open for even more physic processing performance? First, AGEIA has massive amounts of internal bandwidth on the chip. They claim to have nearly two terabits per second (2 Tbits/s) of internal memory bandwidth to work with, many times more than even the fastest processors or GPUs available today. This addresses the needs of our first component of a physic processing system, scale, to a T. Detecting and resolving the collisions of a large number of moving rigid bodies requires this kind of bandwidth to implement the geometric math necessary for the calculations."

but since those numbers havent been posted, I still say your going on BS marketing crap.


Pot, Kettle, Black...
(I will refrain from calling you a liar, you could have been misinformed)

I am not the one been feeding a lot of bogus information.
I am not the one who tried same approch in the PPU forum...just to desert my own thread when facts contradicted the claims...

Terra - Infact I am STILL wondering what this post does in the ATI subforum..I will now report the post as misplaced...no more "hiding" for you...

[Terra]

No they're not. The pipelines generate data which is sent to a location in the GPU's memory. From there, it's not that hard to copy data from the GPU, over the AGP bus, and into system memory to get the results you need. Yes, it stalls a bit when you try to do something like that on D3D, but ATi is developing their physics API such that they're able to talk and interact VERY directly with the GPU, and so the stalling that would normally occur can be easily avoided.

http://www.pcper.com/article.php?aid=225&type=expert. (http://www.pcper.com/article.php?aid=225&type=expert)

The lack of a real write-back method on the GPU is also going to hurt it in the world of physics processing for sure. Since pixel shaders are read-only devices, they can not write back results that would change the state of other objects in the "world", a necessary feature for a solid physics engine on all four counts.

This tidbit is interesting too:
Another interesting issue that AGEIA brought up is that since the Havok FX API, and any API that attempts to run physics code on a GPU, has to map their own code to a Direct3D API using Shader Models then as shader models change, code will be affected. This means that the Havok FX engine will be affected very dramatically every time Microsoft makes changes to D3D and NVIDIA and ATI makes changes in their hardware for D3D changes (ala DX10 for Vista). This might create an unstable development platform for designers that they may wish to avoid and stick with a static API like the one AGEIA has on their PhysX PPU.

And let me go on the record and say that I have NO affiliations with AGEIA...what so ever...
Other than I will be buying a PPU when it's released.

Terra - As I might suspect that "argument" will surface soon....

[serbiaNem]

Lets not forget that depending one how well the API is programmed can make a huge difference to how each method will perform.

I agree with Terra that the gpu will not be a true physics processor, but a sort of after-effect generator. It may have a cool truck explosion, but lets see it handle the player first shooting a crate which collapses, then shooting the pieces multiple times while it's still falling apart.

[Cypher19]

Oh, my, yes. Listen to a hardware enthusiast, and not a graphics programmer.

Here's a summary of how that pixel pipeline data can be obtained:

Step 1) Create default pool render target surface. Using CreateRenderTarget: http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__CreateRenderTarget.asp
(aside: a default pool surface refers to an allocated piece of memory on the video card itself)
Step 2) Create systemmem pool render target surface using CreateRenderTarget: http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__CreateRenderTarget.asp
(aside pt. 2: a systemmem pool surface is a piece of memory that exists in the main system memory which the CPU can easily access)
Step 3) Create pixel shader. Your choice, you can probably use the ID3DXEffect classes or something, or just hte CreatePixelShader function.
Step 4) Apply pixel shader and constants. SetPixelShaderConstant, blah blah blah...
Step 5) Render to default pool render target. Draw*Primitive: http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__DrawPrimitive.asp
Step 6) Call GetRenderTargetData with the default and systemmem RT's as parameters. http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__GetRenderTarGetData.asp
Step 7) Lock the systemmem surface. http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DSurface9__LockRect.asp
Step 8) Get the data from the returned array.


This tidbit is wrong too:
Another interesting issue that AGEIA brought up is that since the Havok FX API, and any API that attempts to run physics code on a GPU, has to map their own code to a Direct3D API using Shader Models then as shader models change, code will be affected. This means that the Havok FX engine will be affected very dramatically every time Microsoft makes changes to D3D and NVIDIA and ATI makes changes in their hardware for D3D changes (ala DX10 for Vista). This might create an unstable development platform for designers that they may wish to avoid and stick with a static API like the one AGEIA has on their PhysX PPU.

Graphics cards maintain a certain degree of backwards compatibility. I can write an SM2.0 shader, and have it compiled as an SM2.0 shader, and run it on a GeForce 6 or 7 series card (i.e. SM3) just fine. The same thing will occur with the Havok FX stuff, and it is under ZERO danger of being "unstable" due to updates to Direct3D.

Now, any questions?

[Xipher]

Havok and partners have stated (I remeber hearing it in the videos, coming from the rep), that they don't intend for them to be able to effect gameplay. It may be possible, but its probably a waste of resources that are better suited for other tasks.

[Josh_B]

Exactly my point.

..looks like Terra has been owned once again.

Assuming there is no caching of meshes or anything.

I can't imagine they would've suggested using the PCI bus unless they knew it wouldn't be a bottleneck. The PCI-E spec has been out for awhile, and if it were an issue, they would've simply waited for the right mobos to become available.

I don't know... that's my $0.02

[Terra]

Oh, my, yes. Listen to a hardware enthusiast, and not a graphics programmer.

Here's a summary of how that pixel pipeline data can be obtained:

Step 1) Create default pool render target surface. Using CreateRenderTarget: http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__CreateRenderTarget.asp
(aside: a default pool surface refers to an allocated piece of memory on the video card itself)
Step 2) Create systemmem pool render target surface using CreateRenderTarget: http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__CreateRenderTarget.asp
(aside pt. 2: a systemmem pool surface is a piece of memory that exists in the main system memory which the CPU can easily access)
Step 3) Create pixel shader. Your choice, you can probably use the ID3DXEffect classes or something, or just hte CreatePixelShader function.
Step 4) Apply pixel shader and constants. SetPixelShaderConstant, blah blah blah...
Step 5) Render to default pool render target. Draw*Primitive: http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__DrawPrimitive.asp
Step 6) Call GetRenderTargetData with the default and systemmem RT's as parameters. http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DDevice9__GetRenderTarGetData.asp
Step 7) Lock the systemmem surface. http://msdn.microsoft.com/library/en-us/directx9_c/IDirect3DSurface9__LockRect.asp
Step 8) Get the data from the returned array.

I am no programmer either, but I do se a LOAD of overhead in what you just described...
But would you call Gabe Newell a programmer?

It’s like you can have a bunch of different things bouncing around so long as they don’t actually touch anything that matters. If you don’t actually have to read the data out – if your AI system ever needed to know about whether or not one of those objects had collided with something else it would run slower by running on the GPU than having it run on the main CPU. So physics that matter is different than physics that makes pretty pictures.





Graphics cards maintain a certain degree of backwards compatibility. I can write an SM2.0 shader, and have it compiled as an SM2.0 shader, and run it on a GeForce 6 or 7 series card (i.e. SM3) just fine. The same thing will occur with the Havok FX stuff, and it is under ZERO danger of being "unstable" due to updates to Direct3D.

Now, any questions?

What does "certain degree of backwards compatibility" mean? :confused:

Either it is or it isn't?

Terra...

[Cypher19]

I am no programmer either, but I do se a LOAD of overhead in what you just described...
But would you call Gabe Newell a programmer?

There's a fair bit of overhead involved with it, yes. But I'd it can be primarily attributed to the D3D relationship with the graphics card. Because ATi's physics system will likely interact fairly directly with the drivers and hardware, it can circumvent the D3D API and get results that get a lot closer to theoretical numbers in more than a few areas, including AGP transfer rates and FLOPS. If you use the D3D API, it will undoubtedly be slow, though, and I believe that's what Gabe is addressing, and also how I think Havok FX works. But mind you, I'm not entirely sure how ATi's physics will work either, it may be pretty stuff there as well (e.g. using render to vertex buffer and such). I was merely saying and demonstrating that it IS possible to get data back from the GPU. Practicality is a seperate issue, though.

What does "certain degree of backwards compatibility" mean?

Well, a 7800GTX doesn't have specialized support for things like SM1.x, SM2.0, SM2.x and SM3.0. Most of the time, the driver will bring 'old' shaders up to the right compiler target and get it working.

[MrNasty]

Cypher19 - wouldn't you need to be running some sort of interpreter to allow the rendertarget thingy data to be seen as physics data?

Also how much memory overhead would that take? Storing physics data in that form would be pretty memory intensive, and storing data in that form for an entire level (as would be required for interactive physics) would surely be difficult.

I know for a fact Ageia's PPU doesn't need much storage at all because it simply computes vectors and tensors and stores the results - and while the transients are pretty large they are dealt with very quickly and so never need to be called to and from memory.

There's no dynamic results handling on a GPU AFAIK, so any intermediate results that couldn't be resolved in a one shader cycle (fluid simulations are one situation I can think of that would need a few passes) would need to be stored externally to the GPU - not good performance-wise, surely?

Another issue - how efficient can ATi's physics model be? Pixel shaders can't talk to each other as they operate, so results from one calculation will have no result on another, unless said calculations are performed repeatedly in light of each others' results - this adds a HUGE amount of memory usage into the balance does it not?

I'm curious, because I know very little about hardware calls and API's, but I know a lot about fluid mechanics and what you need to be able to accurately model continuums etc, in fact I've seen several attempts at what you might call 'hardcore' fluids modelling on GPUs, but these attempts tend to perform either erratically due to lack of communicative parrallelism (the effect I described before) or slowly because of the necessity of storing transient data - hence my questions above.

[Cypher19]

Cypher19 - wouldn't you need to be running some sort of interpreter to allow the rendertarget thingy data to be seen as physics data?


Not really, the returned data would be basically an array of floats that you can use any way you want on the CPU.

Also how much memory overhead would that take? Storing physics data in that form would be pretty memory intensive, and storing data in that form for an entire level (as would be required for interactive physics) would surely be difficult.

It would be, but I couldn't say how great the memory usage would be. It probably wouldn't be that different than any other physics system, but I haven't made a physics system myself, so I couldn't say.

There's no dynamic results handling on a GPU AFAIK, so any intermediate results that couldn't be resolved in a one shader cycle (fluid simulations are one situation I can think of that would need a few passes) would need to be stored externally to the GPU - not good performance-wise, surely?

Nah, there's lot of graphical effects that need multi-pass approaches. One could ping-pong between render targets so that you render some data to one RT, and use that data for another pass to a seperate RT.

Another issue - how efficient can ATi's physics model be? Pixel shaders can't talk to each other as they operate, so results from one calculation will have no result on another, unless said calculations are performed repeatedly in light of each others' results - this adds a HUGE amount of memory usage into the balance does it not?

Getting physics simulation on a pixelshader isn't trivial, and it's not something I can comment on, really. I'm not sure exactly how to calculate physics quickly, let alone how to do it on a pixel shader. I just dived into the thread because Terra was being a bit of a loon.

[Terra]

I just dived into the thread because Terra was being a bit of a loon.

Me?
Are you saying that I am the one 100% wrong in this thread?
That all others are 100% correct...by using fake numbers, GFlops compared across platform/technology, claims about a 3 to1 Physcis prerformance in favour of the GPU over PPU, while totaaly ignoring the architechtual differences like internal bandwith ect.?

Terra - Puuuuuurllleeease.... :D

[Trimlock]

Ever heard of markeeting bullshit before? *L*

marketing bull shit doesn't mean it doesn't work, when you market a product for being able to do something, it should be able to do it, how fast and easy it does it where the bull shit plays into, if they market something that doesn't work or wasn't added in i believe you could say thats false advertisment, so i think both companies have found a way to give us real time physic's aiding at a speed they are happy with, something they are puting money into marketing for

if it is marketing bullshit and ends up being less able then our CPU is at processing physic's and they pumped it up this much, that would be somewhat of a dissapointment, agree?

[Terra]

marketing bull shit doesn't mean it doesn't work, when you market a product for being able to do something, it should be able to do it, how fast and easy it does it where the bull shit plays into, if they market something that doesn't work or wasn't added in i believe you could say thats false advertisment, so i think both companies have found a way to give us real time physic's aiding at a speed they are happy with, something they are puting money into marketing for

if it is marketing bullshit and ends up being less able then our CPU is at processing physic's and they pumped it up this much, that would be somewhat of a dissapointment, agree?

I just watched ATI's video of "physics"
I find it funny that in the end of the video, the mandelbrot end up EXTREMELY pixilated, while the CPU still looks okay
And their R2VB demo gives 22 FPS at 10.000 obejcts, 15 FPS at 20.000 objetcs, 10 FPS at 22.000 objects, 8 FPS at 30.000 objects.

I belive AGEIA has something like 32.000 objects listed as within the specs.
(The ability of the PhysX PPU)
I could be mistaken though...

But as someone else pointed out, the biggest "headache" for NVIDA/ATI is that hte PPU is PCI.
Care to wager how many emtpty PCI slot are out there compared to PCI-E? ;)
It means that current users kan buy a PPU at ~$200-250, no need for a mobo upgrade, as the PPU prolongs the lifespan of a gaming PC.
Hence they talk about "physics" on their produts too, but don't talk to much about that it's effect-physics, not gameplay physics...hence I cal it "marketing BS"

Steal the "spotlight" from AGEIA via PR, not actual preformance.
And I still have to se NVIDA/ATI do anything remotely like eg. the new Cellfactor video.

Terra - Until they step up to plate, that will be my point of view on their solutuions...

[MrNasty]

Interesting stuff Cypher :)

Just to show how different PPU architecture is compared to GPU architecture here are brief rundowns on each:

PPU (http://www.blachford.info/computer/articles/PhysX2.html)

And the best examples of an x1900's architecture

X1900 Overview (http://www.hardwaresecrets.com/fullimage.php?image=3099)
X1900 Quad shader unit (http://www.hardwaresecrets.com/fullimage.php?image=2086)

Right, now let's have a look at the differences:

1. Raw vector power (the most important aspect of modelling physics, speaking as a fluids engineering student having already done a computational fluid dynamics course - wasn't fun, CFX 10.0 users are you with me? :p )

PPU: 16 units capable of 6 floating point executions per clock
16 result dedicated register per unit(all at 32 bit precision)
Total vector ops per clock: 96

GPU: 48 units capable of 1 floating point execution per clock
general memory ring register (precision unknown - ATI quotes 128 bit precision)
Total vector ops per clock: 48 max
Assuming you aren't using them for anything else, like actual pixel shaders for example.

2. Integer-wise

PPU: 16 units capable of 1 integer operation per clock (in addition to the vector calcs)
8 result dedicated register per unit (in addition to the 16 vector ops)
Total Integer ops: 16

GPU: 48 units capable of 1 integer operation per clock- exclusive of vector ops
general memory ring register
Total Integer ops: 48 max, if no vector ops are being done

3. Other differences

PPU: Inter-addressing bus, fully inter-parallel 4-way associativity, no equivalent instruction pipelines (as units can address each other), 1 Dedicated instruction controller per unit, 1 indirect dispatch processor.

GPU: One-way addressing bus, no inter-parallel associativity, 48 instruction pipelines, 1 excecution branch processor per unit (can only control branch instructions), 1 direct dispatch processor (this is where the equivalent functions of the PPU's instruction controller occur on a GPU).

Makes for interesting reading, and while, yes, a GPU's shader unit and a PPU are very similar, they are each much more suited to their task than the other...

For those of you who like quoting Gflops/s bear in mind that ATI/NV quote these with the calculation potential rest of the GPU included (i.e. the bits that do jack for physics processing), hence any comparison with Ageia's PPU is already thrown out the window.

ATI's solution is going to be slow compared to Ageia's, by raw numbers and not even mentioning the architectural execution differences. This analysis also seems to cast doubt on the "but I can run physics AND graphics on 1 GPU" claims made by both ATI and Nvidia - would you like to see your graphics performance cut in half and get 25% of the performance of Ageia's solution (clock for clock)?

All that remains to be seen is how much games use the PPU and whether ATI's (and nvid's) solution will be adequate for the amount of physics utilised.

It's not about raw power.

It's about utilisation and interactivity - and those remain to be seen from either camp at this time.

You can all clap now :D

[Terra]

*claps hands*
But insted of nifty words, I found this "funny" video of a game using PhysX.
Don't know what the game is callled or what the hell is going on...but it looks cool :eek: :D

http://www.gametrailers.com/player.php?id=8913&pl=game&type=wmv

Terra - *Homer mode* Physics...ummmh.....

[BBA]

Interesting stuff Cypher :)

Just to show how different PPU architecture is compared to GPU architecture here are brief rundowns on each:

PPU (http://www.blachford.info/computer/articles/PhysX2.html)

And the best examples of an x1900's architecture

X1900 Overview (http://www.hardwaresecrets.com/fullimage.php?image=3099)
X1900 Quad shader unit (http://www.hardwaresecrets.com/fullimage.php?image=2086)

Right, now let's have a look at the differences:

1. Raw vector power (the most important aspect of modelling physics, speaking as a fluids engineering student having already done a computational fluid dynamics course - wasn't fun, CFX 10.0 users are you with me? :p )

PPU: 16 units capable of 6 floating point executions per clock
16 result dedicated register per unit(all at 32 bit precision)
Total vector ops per clock: 96

GPU: 48 units capable of 1 floating point execution per clock
general memory ring register (precision unknown - ATI quotes 128 bit precision)
Total vector ops per clock: 48 max
Assuming you aren't using them for anything else, like actual pixel shaders for example.



You need to take that a step further:

GPU 650 MHz clock
PPU 100 MHz clock

So, lets say 6.5:1 clock ratio for an XTX:PPU

that gives:
GPU 48 Vector/clock X 6.5 Clock = 312 Vector Ops/ 1 PPU Clock
PPU 96 Vector/Clock X 1 Clock = 98 Vector Ops / 1 PPU Clock.

I'll stop there for now. I am sure you can all extrapolate how thats close to 3:1 adfvantage for the ATi GPU. Now...the nvidia GPU ...wouldn't be much different than the ATi, but nvidia isn't talking about it as much yet.

[BBA]

And let me go on the record and say that I have NO affiliations with AGEIA...what so ever...
Other than I will be buying a PPU when it's released.

Terra - As I might suspect that "argument" will surface soon....

Why wait...buy one right now. (It's been available since 3/15 unless they have all been sold already)



Greetings,



We at AGEIA are excited that you signed up at QuakeCon 2005 expressing interest in our PhysX technology.



PhysX is here now. PC games will never be the same again! This technology will bring exiting new ways to frag your opponent.



We have released 250 AGEIA PhysX cards to Vance Research, a logistics company that handles early card distribution to Modders. They will be sold to extreme modders that can create content and buzz prior to retail availability. These cards will contain two modifiable game-lets based on Unreal Engine 3 and Artificial Reality Engine.



Unreal Engine 3 game-let is the Hanger of Doom demo that we showed at QuakeCon. The Artificial Reality game-let is a multi-player LAN game that uses physics in game play in ways that have never been seen before. AAA games will roll out after retail availability.



AGEIA is planning a contest with a large cash prize for best game-mod. Any content created now will be eligible.



ASUS and BFG manufactured cards are available through Vance Research today @ www.vanceresearch.com . You will need to sign up as a new user. Anticipate two weeks for delivery. Non-disclosure agreement required prior to receipt.



Content can only be publicly shared when cards are available through retail.



Thank you for your continued interest in AGEIA and your contribution to the PC gaming world.





AGEIA Modder Relations

[BBA]

BTW Terra, you mentioned how many spare PCI vs PCI-E slots...right now I have no spare PCI slots, but I have one spare PCI-E X8-16 and 1 spare PCI-E X1 slots (one more X1 slot is covered by the X1900XTX heatsink)

Basically, PCI-E slots are more available as spare slots than PCI slots already, simply because after you install a sound card and a tuner/scsi/whatever, thats how it ends up.

[Xipher]

You need to take that a step further:

GPU 650 MHz clock
PPU 100 MHz clock

So, lets say 6.5:1 clock ratio for an XTX:PPU

that gives:
GPU 48 Vector/clock X 6.5 Clock = 312 Vector Ops/ 1 PPU Clock
PPU 96 Vector/Clock X 1 Clock = 98 Vector Ops / 1 PPU Clock.

I'll stop there for now. I am sure you can all extrapolate how thats close to 3:1 adfvantage for the ATi GPU. Now...the nvidia GPU ...wouldn't be much different than the ATi, but nvidia isn't talking about it as much yet.
So you want to dedicate the entire thing to physics processing, and lose out on all the graphical features it gives you then? I think a lot of the games coming out using Ageias PhysX (like UT2007) are using those vertex and pixel shaders quite extensively. I would guess that would leave little head room for the physics calculations. That is probably WHY Havok FX isn't intended to be used for game play effecting physics, because at the games lower requirements, it probably couldn't even run, and it will still have to dynamically adjust how much of a load it puts on the GPU in case the game wants to throw some more effects on.

[Terra]

You need to take that a step further:

GPU 650 MHz clock
PPU 100 MHz clock

Where the *beeeeeeeeeeeep* have you gotten the idea that the PPU core is 100Mhz? :confused:

So, lets say 6.5:1 clock ratio for an XTX:PPU

Not in this world?

Terra - WTF :confused:

[Terra]

Why wait...buy one right now. (It's been available since 3/15 unless they have all been sold already)

Thanks for nothing( again)

Terra - Fucking PREORDER!!!....

[Terra]

BTW Terra, you mentioned how many spare PCI vs PCI-E slots...right now I have no spare PCI slots, but I have one spare PCI-E X8-16 and 1 spare PCI-E X1 slots (one more X1 slot is covered by the X1900XTX heatsink)

Basically, PCI-E slots are more available as spare slots than PCI slots already, simply because after you install a sound card and a tuner/scsi/whatever, thats how it ends up.

*chough*
You are asuming that the majority of motherboards are PCI-E...
The way ATI/NVIDIA want it too be.
And again, this does not aplly for this world.

Terra - Do you have any affiliation with ATI?!

[BBA]

So you want to dedicate the entire thing to physics processing, and lose out on all the graphical features it gives you then? .


Yes, thats one plan ATi is making..

Thats what ATi is letting you do, you will have one video card for normal 3D, the other for dedicated physics.

Basically, you have todays X1900, it's outdated and you buy a newer model for vista or whatever, and the X1900 becomes sole use for Physics.

Since I upgrade every time a new card comes out, this gives the X1900 a longer life of use.

[BBA]

Where the *beeeeeeeeeeeep* have you gotten the idea that the PPU core is 100Mhz? :confused:



Not in this world?

Terra - WTF :confused:


I got it from the Ageia owner/tech/marketing person at Quakecon.

[Xipher]

Yes, thats one plan ATi is making..

Thats what ATi is letting you do, you will have one video card for normal 3D, the other for dedicated physics.

Basically, you have todays X1900, it's outdated and you buy a newer model for vista or whatever, and the X1900 becomes sole use for Physics.

Eh, seems like a waste of money to me.

[BBA]

Eh, seems like a waste of money to me.


Why? What else will you do with the X1900 after you upgrade? Sell it? I tend to keep older video cards myself, this makes nice use of it.

[BBA]

*chough*
You are asuming that the majority of motherboards are PCI-E...
The way ATI/NVIDIA want it too be.
And again, this does not aplly for this world.

Terra - Do you have any affiliation with ATI?!

Get real man...who in their right mind will buy a PPU if they are not already using a top of the line video card on a PCI-E system board?

You should stop working for Ageia and see what comes out in real life before taking your hard stand.

[BBA]

Thanks for nothing( again)

Terra - Fucking PREORDER!!!....


Im pretty sure I posted this the day it was sent to me...if you missed out, well, who cares anyway. :D

[Xipher]

Sell it, if you have that kind of cash why aren't you running SLI or Cross fire, that second card is in the way. Also if ATi is the only one backing this, then I could care less (Nvidia fan simply due to decent support for Linux).

[BBA]

Sell it, if you have that kind of cash why aren't you running SLI or Cross fire, that second card is in the way. Also if ATi is the only one backing this, then I could care less (Nvidia fan simply due to decent support for Linux).


nvidia is going to support it as well, but only in SLi mode and it will share physics with graphics.

[Trimlock]

its dedicated support for the havocFX engine

[MrNasty]

You need to take that a step further:

GPU 650 MHz clock
PPU 100 MHz clock

Ageia has been keeping its clock speed details under HEAVY wraps - I know, I have a copy of the patent and even that doesn't have full specs and also in a lot of interviews Ageia have said they don't want to give the competition a heads up.

Lets examine what you've spouted anyway:

Power consumption is directly related to fab technique and clock speed.

a 125 million transistor processor dissapating 25 watts. 130nm.

25/125 = 0.2 w/million transistors

a 384 million transistor processor dissipating nearly 100 watts. 90nm.

100/384 = 0.26w/million transistors

ok, so lets see here, based on the fab technique decreasing the power consumption linearly wrt clockspeed (which is almost right - it should reduce by square but never gets there due to inefficiencies):

90/130 * 0.2 w/million transistors = 0.14

so, ATI's clock is 650Mhz which means that Ageia's PPU must be close to:

0.14*(650/0.26) = 350 Mhz

And this is an approximate calculation - but sorry BBA, your BBS is called: 25watts doesn't go nowhere.

so, let's look at your comparison now:

3.5*96= 336

6.5*48= 312

And that, as I mentioned earlier but you so carefully ignored, doesn't take into account that in physics any vector calculations performed usually have to be fed back into each other, something that the dispatch processor will have to handle, and as its already handling the instructions it will be pretty tied up I imagine.

Oh, and who'd you get that number off? a booth babe? :D

[MrNasty]

nvidia is going to support it as well, but only in SLi mode and it will share physics with graphics.

Oh look, another phallacy: nVidia's solutions will no more "share physics with graphics" than ATI's will.

nVidia have also jumped on the hype bandwagon (along with ATI) in saying that the most interesting application of HavokFX will be on a 1 GPU system.

The only difference between their dual-card physics system and ATI's is that with nVid's you have to use 2 identical cards, whereas ATI's (like crossfire) allows you to use any two cards. (ATI gets my vote here :) )

I speculate that ATI is consistently denying you will need to run the cards in "crossfire" mode because they want people to associate crossfire with graphics, in actual fact 2 graphics cards operating in this physics/graphics mode will be indistinguishable from how they work in crossfire, just sans dongle.

nVidia has been far more forthcoming with details than ati have about their "unnamed, unknown, undeveloped" physics API.

[Terra]

Why? What else will you do with the X1900 after you upgrade? Sell it? I tend to keep older video cards myself, this makes nice use of it.

Cause the PhysX is cheaper.
Ebay that X1900, buy a PhysX..and get real physics...

Terra - And even make some money...

[Terra]

I got it from the Ageia owner/tech/marketing person at Quakecon.

I call it BS...

Terra...

[Terra]

Get real man...who in their right mind will buy a PPU if they are not already using a top of the line video card on a PCI-E system board?

I am using a AGP 7800GS(full 7800GT Core)
I would be one of them.
Look at the min requirements dammit..a P4 1.4Ghz(or AMD equvilant)
The PPU dosn't have anything to do with a the GPU

You should stop working for Ageia and see what comes out in real life before taking your hard stand.

1. Why didn't you answer my question?!
I'll repeat then:
Do you have any affiliations with ATI?

2. I know you better than you think:
http://www.hardforum.com/showpost.php?p=1029333760&postcount=49

And let me go on the record and say that I have NO affiliations with AGEIA...what so ever...
Other than I will be buying a PPU when it's released.

Terra - As I might suspect that "argument" will surface soon...."


Terra - Way ahead of you...so why don''t you answer my question now?

[Terra]

And this is an approximate calculation - but sorry BBA, your BBS is called: 25watts doesn't go nowhere.

so, let's look at your comparison now:

3.5*96= 336

6.5*48= 312

And that, as I mentioned earlier but you so carefully ignored, doesn't take into account that in physics any vector calculations performed usually have to be fed back into each other, something that the dispatch processor will have to handle, and as its already handling the instructions it will be pretty tied up I imagine.

Oh, and who'd you get that number off? a booth babe? :D

Nice freaking post! :D

Terra - This is why I wanted the thread out out the ATI subforum...real answers about physics ;)

[Unknown-One]

This is getting insane. We have the Agea PPU for one physics engine, and both ATI and NVIDIA's physics solution for another physics engine...

I wonder if you could have an SLI setup, and both an Agea PhysX card and some random ATi card all in the same system? SLI can work to its fullest because the ATi card will handle one physics engine instead of the second SLI card, and the PhysX card can do its thing in all the games that use it…best of both worlds if you have a motherboard with enough slots for teh two extra physics cards plus whatever other crap you already have in there :p

[BBA]

Oh look, another phallacy: nVidia's solutions will no more "share physics with graphics" than ATI's will.

nVidia have also jumped on the hype bandwagon (along with ATI) in saying that the most interesting application of HavokFX will be on a 1 GPU system.

The only difference between their dual-card physics system and ATI's is that with nVid's you have to use 2 identical cards, whereas ATI's (like crossfire) allows you to use any two cards. (ATI gets my vote here :) )

I speculate that ATI is consistently denying you will need to run the cards in "crossfire" mode because they want people to associate crossfire with graphics, in actual fact 2 graphics cards operating in this physics/graphics mode will be indistinguishable from how they work in crossfire, just sans dongle.

nVidia has been far more forthcoming with details than ati have about their "unnamed, unknown, undeveloped" physics API.


ATi is definitely working on shared physics and GPU processing in single card use...that is straight from ATi when they talked about the Vista driver. I'm pretty sure I read similar about nvidia.

ATi is also going to support dedicated card use for physics in a two card system. nvidia may not, but they will probably do something to counter ATi so you never know.




Your really missing some data there, and stop talking about havok, you guys seem to be getting stuck on something that is really nothing more than an irrelevant side note.

[BBA]

Ageia has been keeping its clock speed details under HEAVY wraps - I know, I have a copy of the patent and even that doesn't have full specs and also in a lot of interviews Ageia have said they don't want to give the competition a heads up.

Lets examine what you've spouted anyway:

Power consumption is directly related to fab technique and clock speed.

a 125 million transistor processor dissapating 25 watts. 130nm.

25/125 = 0.2 w/million transistors

a 384 million transistor processor dissipating nearly 100 watts. 90nm.

100/384 = 0.26w/million transistors

ok, so lets see here, based on the fab technique decreasing the power consumption linearly wrt clockspeed (which is almost right - it should reduce by square but never gets there due to inefficiencies):

90/130 * 0.2 w/million transistors = 0.14

so, ATI's clock is 650Mhz which means that Ageia's PPU must be close to:

0.14*(650/0.26) = 350 Mhz

And this is an approximate calculation - but sorry BBA, your BBS is called: 25watts doesn't go nowhere.

so, let's look at your comparison now:

3.5*96= 336

6.5*48= 312

And that, as I mentioned earlier but you so carefully ignored, doesn't take into account that in physics any vector calculations performed usually have to be fed back into each other, something that the dispatch processor will have to handle, and as its already handling the instructions it will be pretty tied up I imagine.

Oh, and who'd you get that number off? a booth babe? :D

I actually got the number form the Ageia representative. Remember, this was during Quakecon, and a lot has probably changed since the first nda hardware PPU release did not happen until March 15, thats over half a year to ramp up clocks, so you may be right.


Anyway...lets say for argument sake you really know the current clock speed of the PPU, then by your own above statement, the performance difference is what? Worst case it's almost negligable. Don;t you think the GPU would be a better choice since you can still program it for better/different algorithyms as needed, instead of a hardwired PPU?

[BBA]

I call it BS...

Terra...

I would expect no less from you...because your so good at BS yourself. :D

[BBA]

This is getting insane. We have the Agea PPU for one physics engine, and both ATI and NVIDIA's physics solution for another physics engine...

I wonder if you could have an SLI setup, and both an Agea PhysX card and some random ATi card all in the same system? SLI can work to its fullest because the ATi card will handle one physics engine instead of the second SLI card, and the PhysX card can do its thing in all the games that use it…best of both worlds if you have a motherboard with enough slots for teh two extra physics cards plus whatever other crap you already have in there :p


So your saying have two graphics cards plus one for physics AND also add a PPU just to make sure all angles are covered?

Sounds good to me. Of course motherboard layout might complicate that a little but if the PPU goes PCI-E it might actually be workable.

[Xipher]

Don;t you think the GPU would be a better choice since you can still program it for better/different algorithyms as needed, instead of a hardwired PPU?
NO! Doing software calculations is ALWAYS slower, and requires a faster clock just to catch up with hardware acceleration. I have said it multiple times, the GPU will more then likely need MULTIPLE clock cycles to do the same calculation a chip dedicated to physics processing can do in one, simply because its a hardware instruction. Every thing you gain with a faster clock is lost when you need to do 3 instructions just to do what a PPU does in 1. This is why we have 3D acceleration, because hardware acceleration is faster then having the CPU do rendering, even when the clock speeds are so different.

[quadnad]

YOUR: The possessive form of you. "Your boots are in the hallway."

YOU'RE: Contraction of you are. "You're going down to the store today?"

I just had to.

[Low Roller]

ATI's GPU physics is Havok FX, not something different. There are two competing physics API's right now, Havok FX and Ageia's PhysX(formally Novodex).

Havok FX is not used for gameplay physics at all, and it costs the developer an additional licesnse fee to use over and above Havoks regular physics products.(which use the CPU)

From the Havok FX website:So, for acceleration of game-play physics, along with the core game-logic, special purpose physics hardware is not the answer. Multi-core CPU architectures are the path to speed.

For effects physics, however, the GPU offers the most compelling and sustainable promise to added visual fidelity in games. Relative to proprietary hardware, GPUs also have a clear advantage as a pre-existing technology familiar and readily available to consumers and game developers, providing other benefits such as wide-spread availability, commodity pricing, and mature standards for hardware and software interfaces.

What Is The Difference Between Game-Play Physics And Effects physics?

Game-Play Physics affect how a game is played from moment-to-moment, and is generally computed on a computer’s central processing unit (CPU). Physical changes that you cause in the game or that happen to you or around you– like knocking over a box, and then climbing up on it - change what you may want to do in each instant of game play. Both game-play physics and game logic demand instant access and tolerate no detectible latency to preserve the game-play experience. The close proximity between physics, game logic, and memory, defines game play and generally demands that these systems execute together on a game system’s CPU. Effects physics is an emerging domain that promises to deliver an increasing array of visually impressive effects that are based on physical principles – but which place far fewer demands on the game’s logic.

Effects physics – a close cousin to visual effects now computed on GPUs – add to the visual complexity of a game and help increase a player’s immersive experience. As visual phenomena, effects physics need to be convincingly real but do not profoundly affect game play. They can merely fill in the player’s view of the game, creating a richer, more convincing environment- but may not affect the choices a player can make from moment-to-moment.

Does Havok FX Cost Extra?

Yes. Havok FX is not part of any existing product or product bundle. It is an optional add-on product that game developers will be able to license for their platform or console based games.

Will Havok FX Be Included As A Part Of Havok Complete™, Havok Physics™ Or Havok Complete XS™?

No. Havok FX will be an optional product than may be licensed separately, and used in conjunction with any of Havok’s products and product bundles.The developer not only has to shell out cash for a regular Havok Physics license(CPU physics), they also are forced to purchase a Havok FX licesnse(GPU physics) seperately.

In contrast, if the developer chooses to use Ageia's PhysX software, they get both effects and gameplay licenses for free.(as long as they support the PPU)

How many upcoming games will be using Havok FX? As of now, zero.

[Terra]

I would expect no less from you...because your so good at BS yourself. :D

I take it that you are affilitated with ATI, since you won't answer my question about it.
I seem to remember that there are some rule about that if you are affiliated with companies, you have to make that clear in your profile.
Or be banned/deleted...

Terra - Am i mistaken?

[Terra]

ATI's GPU physics is Havok FX, not something different. There are two competing physics API's right now, Havok FX and Ageia's PhysX(formally Novodex).

Havok FX is not used for gameplay physics at all, and it costs the developer an additional licesnse fee to use over and above Havoks regular physics products.(which use the CPU)

From the Havok FX website:The developer not only has to shell out cash for a regular Havok Physics license(CPU physics), they also are forced to purchase a Havok FX licesnse(GPU physics) seperately.

In contrast, if the developer chooses to use Ageia's PhysX software, they get both effects and gameplay licenses for free.(as long as they support the PPU)

How many upcoming will be using Havok FX? As of now, zero.

Well, take in considaration that BBA thinks AGEIA is not a hardware company, but only out to sell software licenses, why do you think he would believe you?:
http://www.hardforum.com/showthread.php?t=1037928

Ageia is NOT a hardware company

Terra...

[Terra]

I actually got the number form the Ageia representative. Remember, this was during Quakecon, and a lot has probably changed since the first nda hardware PPU release did not happen until March 15, thats over half a year to ramp up clocks, so you may be right.

So we can take your post and ignore it as another exsample of you using fictional numbers? :)


Anyway...lets say for argument sake you really know the current clock speed of the PPU, then by your own above statement, the performance difference is what? Worst case it's almost negligable. Don;t you think the GPU would be a better choice since you can still program it for better/different algorithyms as needed, instead of a hardwired PPU?

You my friend are very misinformed.
The PPU dosn't have it's functions "hardwired" so they can't be changed/programmed.
Let's look what they are doing with the drivers:
http://www.ageia.com/developers/downloads.html

What’s New in v2.4:

Acceleration of jointed rigid bodies on AGEIA PhysX processor
Additional fluid collision types (spheres/capsules) accelerated on AGEIA PhysX processor
Hardware Scene Manager for the AGEIA PhysX processor with automatic connection between software and hardware scenes, particularly for instantiating particle fluids.
Heightfields support in AGEIA PhysX software runtime engine, including new shape type mainly for terrain, offers reduced memory consumption and increased speed.
User control of threads [cross-platform]
Common interface to support multithreading on different platforms, which gives developers more flexible control over the threading performed by the SDK
Enhanced collision detection support in AGEIA PhysX software runtime engine
Collision detection by dynamic versus dynamic rigid bodies
Visual remote debugger (VRD) [cross-platform]
Development tool to visualize physics data when optimizing game titles
PML import/export layer in the SDK build component
A common library for import/export included as part of the SDK build component. Supports binary and COLLADA physics formats


Or this:
http://www.pcper.com/article.php?aid=225&type=expert&pid=3

In fact, they have already seen the opposite occurring; upgraded firmware and drivers for the PPU has opened up new features and options to programmers on the same hardware. Cloth simulation is the first example of this as it wasn’t ready initially when the SDK went out to software developers but has since been perfected and added into physics engine, without a need for a hardware upgrade.

So the answer to you very misguided question must be:
No the GPU will not be better/faster for real physics!


Terra - And the saga continues...

[ikarinokami]

I find this whole thing rather odd, if only because the last million or so years of evolution in biology or the last 70 or so years in micrprocessor or the last 10,000 years in human warfare has taught us anything it's that specialization will always make you better at task as opposed to general use. Why is this even a debate? We have one processor that has been designed from the ground up to do one function. It's like BBA has never heard of the DSP. Specialized chips will invarible be faster at the functions that they have been designed for as opposed to ones that have been disigned for general task or adapted from one task to another. What is even more troubleing about his mertiless arguement is that it ignores the most important point. The whole point of physics is to ulitmately change the gameplay, thats why it's a big deal, Since the return of the analog stick there hasn't been anything new that has allowed games to really go new places, gameplay wise. Not only is the ATI/Nvidia solution not a efficient it doesn't do the physics we really want them to do in games. I can only hope people see this red herring for it is and desist. The PPU needs to be supported so we can finally move beyond the pretty lights.

[Terra]

I find this whole thing rather odd, if only because the last million or so years of evolution in biology or the last 70 or so years in micrprocessor or the last 10,000 years in human warfare has taught us anything it's that specialization will always make you better at task as opposed to general use. Why is this even a debate? We have one processor that has been designed from the ground up to do one function. It's like BBA has never heard of the DSP. Specialized chips will invarible be faster at the functions that they have been designed for as opposed to ones that have been disigned for general task or adapted from one task to another. What is even more troubleing about his mertiless arguement is that it ignores the most important point. The whole point of physics is to ulitmately change the gameplay, thats why it's a big deal, Since the return of the analog stick there hasn't been anything new that has allowed games to really go new places, gameplay wise. Not only is the ATI/Nvidia solution not a efficient it doesn't do the physics we really want them to do in games. I can only hope people see this red herring for it is and desist. The PPU needs to be supported so we can finally move beyond the pretty lights.

And this is why peoples "rank" mean squat.
Great post, hit right on the nail :)

Terra - Welcome to [H]ard, ikarinokami :)

[BBA]

NO! Doing software calculations is ALWAYS slower, and requires a faster clock just to catch up with hardware acceleration. I have said it multiple times, the GPU will more then likely need MULTIPLE clock cycles to do the same calculation a chip dedicated to physics processing can do in one, simply because its a hardware instruction. Every thing you gain with a faster clock is lost when you need to do 3 instructions just to do what a PPU does in 1. This is why we have 3D acceleration, because hardware acceleration is faster then having the CPU do rendering, even when the clock speeds are so different.

The whole last sentence of your argument here negates the argument when you consider that 3D acceleration is done by reprogramming the GPU in the same fashion. In your comparison, you would say that all 3D in a GPU is handled in software.

My point is since the GPU is not hardwired and the PPU is, the PPU can become obsolete.

It only makes sense because if hardwired non programmable 3D processing was so much better, then the GPU's would not be programmable...and they have been programmable since the days of GForce 1 for one reason: flexibility. The PPU has none.

[BBA]

I find this whole thing rather odd, if only because the last million or so years of evolution in biology or the last 70 or so years in micrprocessor or the last 10,000 years in human warfare has taught us anything it's that specialization will always make you better at task as opposed to general use. Why is this even a debate? We have one processor that has been designed from the ground up to do one function. It's like BBA has never heard of the DSP. Specialized chips will invarible be faster at the functions that they have been designed for as opposed to ones that have been disigned for general task or adapted from one task to another.

Not true. Specialized one purpose chips are only made for scale of economy. They may be more efficient as the ONE thing they do...but the fastest chips are typically more advanced and programmable in todays reality. Besides, in nature, only the adaptable survive, evolution has made many creatures extremely adaptable and many that were not have perished (the last figure I read was over 90% of all species that have existed on earth are now extinct, mostly as they could not adapt to the changing world.)


You need to read the above response from me as well. Your evolution does not let you understand reality.



What is even more troubleing about his mertiless arguement is that it ignores the most important point. The whole point of physics is to ulitmately change the gameplay, thats why it's a big deal, Since the return of the analog stick there hasn't been anything new that has allowed games to really go new places, gameplay wise. Not only is the ATI/Nvidia solution not a efficient it doesn't do the physics we really want them to do in games. I can only hope people see this red herring for it is and desist. The PPU needs to be supported so we can finally move beyond the pretty lights.

The whole point of physics is one of two things: Change gameplay or increase visual effects, and you can limit the degree of both in either case. There will be no successful multiplayer games with physics that have different gameplay with or without the PPU just because of alienation of non PPU customers but, there may very well be single player games that depend on PPU for certain types of gameplay.

[Terra]

The whole last sentence of your argument here negates the argument when you consider that 3D acceleration is done by reprogramming the GPU in the same fashion. In your comparison, you would say that all 3D in a GPU is handled in software.

My point is since the GPU is not hardwired and the PPU is, the PPU can become obsolete.

It only makes sense because if hardwired non programmable 3D processing was so much better, then the GPU's would not be programmable...and they have been programmable since the days of GForce 1 for one reason: flexibility. The PPU has none.

Again, the PPU is not "hardwired" so that new functions&features can't be added!
STOP LYING!

Terra...

[BBA]

Again, the PPU is not "hardwired" so that new functions&features can't added!
STOP LYING!

Terra...


So now your saying the GPU is capable of being programmed to do more than just physics? Blows most of your arguments.

Or are you saying it is advanced and programmable like a GPU?

[BBA]

I'm sorry to say, I do not have any desire to continue wasting my valuable time in this pointless thread Terra has taken so far off reality.

Believe what you want...it will not matter to me, so have fun.

[Terra]

So now your saying the GPU is capable of being programmed to do more than just physics? Blows most of your arguments.

Or are you saying it is advanced and programmable like a GPU?

You answer me if you are affiliated with ATI or not and I will give you an answer ;)

The [H]ard|FORUM Rules (http://www.hardforum.com/showthread.php?t=760666)

(12) Do not IMPERSONATE other individuals or falsely represent yourself.

(19) ADVERTISING, site pimping, contests or any type of business promotion is not permitted. Soliciting for fraternal organizations, humanitarian causes or personal rewards is prohibited unless approved prior to posting.

And again, your earlier statments about the PPU being "hardwired" and unable to get new features ect. via driver/firmware updates was still a flat out lie!

Terra...

[MrNasty]

See, now it's reached the stage of arguing for the sake of arguing.

BBA - you need to read up more on the definition of "hard-wired" and "programmable" in relation to this topic, in addition to what makes a micro-architecture good at what it does. I wouldn't want you to be ridiculed or plain ill-read up any more than you already have been, but it's your call. I just glad you don't make chips - they'd be a mess. :(

Please leave that "I know it's a 100Mhz part" off the board - I've googled it and any such speed rating, were it given out freely, would have been shithot news right now, yet nothing's around. Sorry dude but that was bull or you were mislead. Also your explanation of nvidia's physics solution was incorrect and your assertion that PPU's are not programmable was incorrect too - PPU's are programmable straight through Ageia's PhysX API, while GPUs pixel shading units are programmable through shader programs interpreted by the D3D API or, in ATI's case, an as yet unconfirmed/undefined API - bear in mind this last bit of info as it is the crux of this discussion: can ATI pull off interactive physics through their GPU? If so it will have to be through ATI's own API.

Terra - stop trying to report him, I doubt he works for ATI - He seems to be a deluded ATI fan that has taken 1 article from an ATI representative as gospel. Sad but true. And don't back yourself into a corner for the sake of a rebuttal :p

The article BBA takes a lot of this from also states that Havok on an nVidia 7900GTX runs about 5% of the speed of an X1900XTX - I think it's safe to call the info in this article marketing bumf.

Let's get this discussion back OT - Cypher had some really intriguing points earlier and there were some goodies made elsewhere too (BBA, yes, you too - earlier).

Fact: We have hard performance numbers for ATI's part and (thanks to me thankyeverymuch ;) ) some approximate performance numbers for Ageia's part. Incidentally that calculation for clockspeed shouldn't be too far off.

The problem is these numbers aren't a great comparison. Not only will ATI use a physics API that is completely different to Ageia's, but the architecture each is using is substantially different.

What is funny about this situation is that there is a huge amount known about the PhysX API but little about the PPU, and nothing known about the ATI API and lots known about the ATI hardware.

Coming back to this - edit later.

[Xipher]

The whole last sentence of your argument here negates the argument when you consider that 3D acceleration is done by reprogramming the GPU in the same fashion. In your comparison, you would say that all 3D in a GPU is handled in software.

My point is since the GPU is not hardwired and the PPU is, the PPU can become obsolete.

It only makes sense because if hardwired non programmable 3D processing was so much better, then the GPU's would not be programmable...and they have been programmable since the days of GForce 1 for one reason: flexibility. The PPU has none.

That's the thing, what I am talking about is the HARDWARE INSTRUCTIONS are made to handle graphical manipulation calculations for a GPU. I'm not saying one OpenGL or D3D API call is a single instruction, but the instructions that ARE implemented in hardware revolve around the kind of calculations for graphics manipulation. The PPU is similar, it just has them oriented for physics. They can add more features later on, using the instruction set currently included, and if needed, use the CPU for certain calculations if something is missing (but they could add that to the next gen card, and remap the API call too the new hardware accelerated version, and the game wouldn't need to change a thing too use it :))

The GPU Vertex and Pixel shaders, and the PPU arch are probably similar, but its the instruction set differences, and the fact it has dedicated memory that won't be taken up by texture and geometry data it can use. Don't forget, current gen games are using massive textures and high polygon models too look so good, that isn't exactly small data wise. Doom 3 can fill up a 256MB card on ultra (no compression for the textures) and make it stutter (recommended minimum for ultra is 512MB video card).

[ikarinokami]

double post. where is the delete key

[ikarinokami]

The whole of the PPU is to specialized in a single task. It's cheaper thats its not advanced as a GPU is meaningless all it has do is be better at doing physics.. This is the whole point of specialization. It does not need to do alot of different things, it only needs to what it does well.

I believe the cellfactor disagree with this propostion. currently their demo is only possible on a LAN connection, but technology moves foward. As for alienation, that what they said when 3D required accelarator games came out.The same can be said of flight simulator games that really only work well with flight sticks. Progress is not always good, but thats not the case.

I am still puzzled why you would prefer the limited and inferior technology of ATI/Nvidia, that don't bring anything new to the table. It's purely technology for the sake of technology. I could understand if this were at least a battle of equals but its not, other than the comfort of familiarity that nvidia or ATI offer I don't see any net gain in adopting either of these technologies.

P.S. you are not applying evolution correctly, but your not to be faulted for that, most people outside the biological sciences are very misinformed about the mechanics of evolutional thoery.

[MrGuvernment]

Yup. This is why you don't have CPUs handle game graphics any more, despite the fact that your average CPU is lightyears more powerful than your average GPU. :D

Um, i hardly think a CPU is light years ahead of a GPU.... ? Why do you think that - if that is the case why are CPU's often the bottleneck for higher resolutions? And why arent developers putting more work on the CPU if it is so much better then a GPU....

[serbiaNem]

I am really interested in if BBA is ATI affiliated or just trying to make people think he is by not answering on that question.

My take on the GPU vs. PPU battle boils down to which one brings something new to the table. The GPU method will be used just for eye candy physics, while the PPU will be used for eyecandy/gameplay physics. What would you rather have? An exploding object sending out bits of useless debris everywhere, or an exploding object sending out bits of environment/enemy destroying debris everywhere? Also, who wants to waste an expensive videocard on eyecandy physics when you can buy a $200 ppu?

[snefan]

this "ati physics" is still on the 'idea board' in ati's office. and thats a fact

[mrjminer]

But isn't everything just eye candy? I mean, isn't the whole premise behind new graphics cards just the introduction of more and better eye candy? Otherwise, we'd all be playing Ultima and Doom 1 non-stop. Realistically, I don't see how the ATI solution would fail in giving more candy and being cheaper, as you can toss your old card in or, by that time, buy an x1600 for less than a PPU.

So, with a GPU to do physics I get a better looking game than with a PPU while software can sufficiently direct the card to render physic-like effects--for less money? Count me in for this, if it works. Just think about it, all the PPU is doing is helping direct the graphics card on what to render. If software + hardware can create the same effect, then they are, effectively, PPU's, despite whether or not the physics are "real."

[Terra]

But isn't everything just eye candy? I mean, isn't the whole premise behind new graphics cards just the introduction of more and better eye candy? Otherwise, we'd all be playing Ultima and Doom 1 non-stop. Realistically, I don't see how the ATI solution would fail in giving more candy and being cheaper, as you can toss your old card in or, by that time, buy an x1600 for less than a PPU.

Becuase it only adds "eye candy"..not all the sweet stuff of real gameplay physics.
And no the PPU is about a jella lot more than "eyecandy"...

So, with a GPU to do physics I get a better looking game than with a PPU while software can sufficiently direct the card to render physic-like effects--for less money? Count me in for this, if it works. Just think about it, all the PPU is doing is helping direct the graphics card on what to render. If software + hardware can create the same effect, then they are, effectively, PPU's, despite whether or not the physics are "real."

Someone here claimed the PPU would get creamed by GPU's....but ran off.
And don't even get me started on how much FASTER the PPU are at physics than a CPU.
Try looking around at this page:
http://personal.inet.fi/atk/kjh2348fs/ageia_physx.html


Terra...

[freddiepm61]

using graphics for physics is dum.
why?
because it will be using up power that is meant to go on graphics, so you will have to turn your settings down.
moreover, you will have to turn settings down even further as the extra eye candy (as this is all it can be as there can be no collision detection if it is done by the gpu) will require more graphics horsepower to show all of the rocks and smoke etc.

therefore doing it via the gpu is dum!!!
however at the moment as is buying a ppu. i would wait till second gen to see benefits etc. as with most things the true early adopters usually get a bit screwed. of course if I had the cash I would definately buy one!
f

[mrjminer]

using graphics for physics is dum.
why?
because it will be using up power that is meant to go on graphics, so you will have to turn your settings down.
moreover, you will have to turn settings down even further as the extra eye candy (as this is all it can be as there can be no collision detection if it is done by the gpu) will require more graphics horsepower to show all of the rocks and smoke etc.

therefore doing it via the gpu is dum!!!
however at the moment as is buying a ppu. i would wait till second gen to see benefits etc. as with most things the true early adopters usually get a bit screwed. of course if I had the cash I would definately buy one!
f

Yah, you bring up some good points about having to lower settings and the such. Also, I suppose that quite an awesome power supply would be needed to run two cards in Xfire + an gpu to do physics, which would cost money to upgrade. Personally, I'm probably going to do what many are doing and just run a DX10 card with a former ATI card for PPU.

Anyways, I guess we'll all see what the benchmarks bring when the come out, but I'm probably going to have to go this route anyway because I couldn't afford to buy a PPU if it turns out to be better :rolleyes:

[$BangforThe$]

Interesting thread to say the least.

After reading the whole thing. I have one question that keeps poping into my mind.

How is the new Vista API going to be incorperated into this PPU thing?

If i was a betting man . I would say that ATI's and Nv's approach to PPU is the way to go as Vista API is going to adderess all the games code including pyhics and treat it all the same . No more optimizations for either ATI or Nv. I don't know how this will affect PPU.

The only thing I am sure of is that ATI is working very closely with MS. So I will put my money on ATI.

[$BangforThe$]

Yah, you bring up some good points about having to lower settings and the such. Also, I suppose that quite an awesome power supply would be needed to run two cards in Xfire + an gpu to do physics, which would cost money to upgrade. Personally, I'm probably going to do what many are doing and just run a DX10 card with a former ATI card for PPU.

Anyways, I guess we'll all see what the benchmarks bring when the come out, but I'm probably going to have to go this route anyway because I couldn't afford to buy a PPU if it turns out to be better :rolleyes:


If what I read is correct about DX10 . I really don't see how the none DX 10 card is going to take instructions from Vista API using DX10. DX 10 requires unified shaders

How and if it can offload pyhsics to dx9card remains to be seen . Just a thought.

[MrGuvernment]

Interesting thread to say the least.

After reading the whole thing. I have one question that keeps poping into my mind.

How is the new Vista API going to be incorperated into this PPU thing?

If i was a betting man . I would say that ATI's and Nv's approach to PPU is the way to go as Vista API is going to adderess all the games code including pyhics and treat it all the same . No more optimizations for either ATI or Nv. I don't know how this will affect PPU.

The only thing I am sure of is that ATI is working very closely with MS. So I will put my money on ATI.


I dont think Vista the GUI will use a PPU because it is not like when you click my computer it explodes into many pieces and will affect if you can click something else or not....

[Mitolo]

OMG wow I have to print this thread out as it is only getting WARM and soon we will all know the truth ;)