• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

My Test of GRAW running with and without physX

ihatelag said:
Awesome, cheers for the link!

welcome...

Upon reading the article again, I thought I'd just add a personal observation - it's fine to slate the game for a poor implementation of PhysX, however it is not fair to slate Ageia for something they have no real control over (ignoring the possibly poorly thought out release timing of the hardware).

It's new technology, and as we have seen in the development of PS2 and XBOX games over the years, it is as much about learning to correctly implement the hardware and it's capabilities as it is about the hardware producers supplying a sufficiently comprehensive development kit.

It's strikes me as important to try to maintain a balanced view on our opinions and filter out the "we're better than you" pap that the companies are going to use as part of their marketing and PR strategy.
 
I agree, you have to be fair to AGEIA, it is still early days...which is why I am waiting for the second/third generation release of their PPU before I even consider purchasing one.

Pushing aside for the moment, my comments from my previous posts about bandwidth and communication overheads, I wonder if AGEIA's NovodeX SDK was to blame for the poor quality implementation? Wasn't it scalable enough to implement in GRAW properly?

I wonder exactly how much support and communication the developers of GRAW got from AGEIA during the developement?

It probably would be more embarrasing for AGEIA if we decided to email the developers of GRAW, and ask if they have utilised the NovodeX SDK properly...I would bet any money that they would say AGEIA supported them to their fullest extent...which would be funny considering the poor results from lots of people using the PPU.
 
I'm still of the opinion that this is a chicken & egg problem.

If it requires loads of work on the part of the developers to get PhysX integrated with existing Havoc-powered titles, then what's the incentive to buy it? Especially when benchmarks are showing it actually slowing down the game.

Developers aren't going to do away with Havoc because what then would non-PhysX people have? No physics at all? PhysX is only going to augment what's already there in terms of physics calculations - and if it can't offload Havoc work onto its own architecture then its doomed imo.

I'm not trying to be all doom mongering, but Havoc is a tried-and-tested implementation that is not hardware specific, as such it's obviously going to be the first choice when it comes to physics implementations by game developers.

Thinking that games companies would write games that only operate properly on certain hardware configuration is lunacy. There's a World of difference between them adding support for things like Creative EAX, Sidewinder joysticks, etc which don't detract from the actual core game....
 
Utterly agree on both counts - any judgement formed on the PhysX currently is going to be based on the implimentation by third parties, for which we will never really know the extent of communication between them and Ageia. You would hope and I would suspect that Ageia helped them to the fullest extent, and if this is the case, then it doesn't bode well for the PhysX initial iteration.

Like you said Durzel, it is something of a 'catch 22' - Ageia wants to gain a share of the gaming market, but the gaming market caters to the largest faction of consumers.
 
[quote='The inquirer"]
"I think the main problem is still the content and along with the fact that people still expect too much from it." [/quote]

Yes..we expect too much...we expected the PPU to work...I guess we asked for too much :)
 
"I think the main problem is still the content and along with the fact that people still expect too much from it."

Yes... wanting a £200 card to do more than display chunks of cheapo looking debris (though this is more of a game related issue, rather than the hardware itself, but it's all linked) and to increase framerates is asking too much :eek:
 
I noticed in the statement from Havok that "Havok’s logo is on the GRAW PC box, substantiating Havok’s use in the game (confirmed by Ubisoft marketing)."

So is GRAW built from the ground up using Havok SDK for all physics/collision detection etc but then switches over to Physx SDK for these specific particle affects? Perhaps if the game is running two Physics engines simultaneously and changing between them it might explain some of the performance issues we are seeing.
 
I have refrained from calling Physx capabilities useless, I think the educated dissapointment is the lack of timing and control of the product at launch, perhaps one could call it marketing infancy?
 
Dutch Guy said:

Look's like Anadtech also experienced a drop in frames per second in GRAW, and reghardware shares the same sort of feeling as ppl in this thread and thinks that this is early days in the PPU world (which I agree) but currently the price of the PPU doesn't justify the results from the PPU.

Reghardware at the time of writing said that the PPU will only work in 32 bit versions of windows? Does it work for 64 bit now?
 
G.O.A.T said:
I noticed in the statement from Havok that "Havok’s logo is on the GRAW PC box, substantiating Havok’s use in the game (confirmed by Ubisoft marketing)."

So is GRAW built from the ground up using Havok SDK for all physics/collision detection etc but then switches over to Physx SDK for these specific particle affects? Perhaps if the game is running two Physics engines simultaneously and changing between them it might explain some of the performance issues we are seeing.


Yes, I dont understand why no one else has picked up on this. Even though the PPU is present, they only use it for particles, and keep havok for everything else - it's stupid sloppy design. If a PPU is present, ALL physics should be moved onto it. I really do feel for Ageia here, because instead of showing the power of the card GRAW just utilises it in an inefficient way. The slowdown could probably be put down to overheads of using a full-on PPU for deciding where some particles fall.
 
Back
Top Bottom