Full Detail Response from AGEIA

[aldo]

I think it's highly unlikely they haven't benchmarked their own software on the platforms they're providing for. I mean it'd be virtually impossible not to. They're simply telling us that because they don't want us to ask for that data or to be seen to be withholding it - because they now realise, probably after some sticky conversations with various people, that explicit performance comparison isn't very diplomatic.

Nothing could be more obvious. You nailed it.

 

[Titanio]

I'm not getting on you, but I have to ask you a question. Do you really believe that? I don't know what to believe now. Seriously.

Err, yes. Look around. We have AGEIA slides, for goodness sake, that make statements about relative performance (if one's translation is to be believed). We have SenatorMonkey here reporting their discussion about "differences in suitability", if you want to put it in soft terms, between fluid/cloth on Cell and X360. It's evident these presentations has caused trouble for AGEIA with certain people and they're now putting distance between themselves and those statements.

 

[Xenus]

The implied difference in performance was from assumptions about the number of compute elements and memory architectures, and how well these might fit to various simulation algorithms as enabled by typical game developers.

I never thought there was any confusion about this. It was pretty clearly stated the first time that the 360 limitation was due to lack of resources.

You don't get PR speak do you earlier they stated that it was not due to publicly released information. It means they leaked something they shouldn't have and are now trying to cover there asses.

 

[seismologist]

You don't get PR speak do you earlier they stated that it was not due to publicly released information. It means they leaked something they shouldn't have and are now trying to cover there asses.

It could be that, but I'm guessing the statements were made based on assumptions about the architecture. The ******s pinned them into a corner until they finally broke down to admit that the statement was not based on any hard data (Even though the conclusions based on the architectures are pretty obvious).

 

[fireshot]

Titano you are trying too hard.

 

[Xenus]

If you read the email it says its not based on "any public data" which is much different than "any data".

 

[Titanio]

Titano you are trying too hard.

Take from it what you will, if you wish to turn a blind eye to what was said previously feel free. The main thing that was evidently inaccurate was that X360's SDK would not include fluid simulation, but the rest..?

I think what was said in the looser environment of a dev conference might be a little more insightful than a reactionary PR statement, so I'm not going to just ignore what was said previously, particularly in terms of direct quotes or AGEIA's own slides for example. If they can provide more information than we were meant to get, so be it, I'll take all the info I can.

 

[onetimeposter]

It could be that, but I'm guessing the statements were made based on assumptions about the architecture. The ******s pinned them into a corner until they finally broke down to admit that the statement was not based on any hard data (Even though the conclusions based on the architectures are pretty obvious).

are you saying i broke them down

 

[onetimeposter]

Take from it what you will, if you wish to turn a blind eye to what was said previously feel free. The main thing that was evidently inaccurate was that X360's SDK would not include fluid simulation, but the rest..?

I think what was said in the looser environment of a dev conference might be a little more insightful than a reactionary PR statement, so I'm not going to just ignore what was said previously, particularly in terms of direct quotes or AGEIA's own slides for example.

Titanio I just replied responding with critical comments from my side that you made like direct qoutes and AGEIA's own slides so I'll see what the response is

 

[mckmas8808]

So more than likely they are trying to keep relations between them and their partners happy? I guess will see more in the future.

 

[Titanio]

Titanio I just replied responding with critical comments from my side that you made like direct qoutes and AGEIA's own slides so I'll see what the response is

I queried them further and they said "AGEIA also will not be benchmarking console platforms since the SONY and Microsoft platforms are not our products and we will not represent their performance".

It honestly does sound to me like they got a slap on the wrists, but again, feel free to take your own conclusions from this.

 

[mckmas8808]

I queried them further and they said "AGEIA also will not be benchmarking console platforms since the SONY and Microsoft platforms are not our products and we will not represent their performance".

It honestly does sound to me like they got a slap on the wrists, but again, feel free to take your own conclusions from this.

See now I'm starting to believe you and what you're trying to say. Why would they announce something important as they did to developers that actually make games for god's sake, yet tell the truth to some forum posters on some website? Aren't they trying to make millions of dollars from this tech? Wouldn't they want x360 devs to be excited about what they have? Why would they tell them one thing and say another to some guys?

It is starting to look like they got a good slap and are making those partners happy by reporting something different. Man what's going on?

 

[Titanio]

It is starting to look like they got a good slap and are making those partners happy by reporting something different. Man what's going on?

Just take what you can get when you can get it, because from here on out you can be sure they'll have their tech people on a very short leash with regard to what they can and can't say about performance across the systems. The CEDEC/GDC presentations provide an insight in that regard that we're very unlikely to get again without explicit approval from MS and Sony, which will never happen because one of them is bound to come off worst than the other.

It's unfortunate, but there ya go.

 

[BTOA]

So in the end, Xenon will still take a big beating while Cell okay when both are running Ageia's physic engine?

 

[hongcho]

Maybe it's not a "puiblic" data because they really haven't optimized the code enough on the console platforms. I am sure they've run their existing code to get some numbers.

Hong.

 

[Titanio]

So in the end, Xenon will still take a big beating while Cell okay when both are running Ageia's physic engine?

They indicated at the conferences that PS3 had more capacity/potential when running novodex than X360 and/or that it had a more suitable architecture for fluid/cloth dynamics. I'm sure you can do enough physics on either chip to bring both to their knees, the amount required probably differs. The official PR line now is that any suggestions made at these conferences was prediction/expectation based on architectural/system analysis rather than benchmark data. So..draw your own conclusions I guess.

 

[seismologist]

are you saying i broke them down

yes, but I'm not sure what you're trying to prove though. They may not have actual data, but that still doesn't make Xenon any better at handling fluid dynamics.

 

[scooby_dooby]

So the CELL CPU has more potential for physics?

In related news...water has been confirmed to be wet.

 

[one]

Of course they are supposed to be doing "educated guesses" at those developer conferences, do you know even the EULA of Visual Studio state that you can't publish any benchmarking data by yourself IIRC... It's just like DeanoC did in his presentation for GDCE only with publicly available information. How you interepret them is basically free.

 

[scooby_dooby]

This is soemthing everyone expecte to be the case, so what's so interesting?

The only "new" info to be discussed was that the SDK for x360 would not contain FD, that's been confirmed to be bunk.

So we're left with what? CELL has more physics potential than XeCPU, what else is new?