• 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.

Cellfactor and Physx

XtAsY said:
The PhysX card I believe will take off with the arrival of Unreal Engine 3 based games as the PhysX API is coded directly into the engine. When games like Unreal Tournament 3 and Bioshock arrive, I'll probably get one if the differences are worth it.

Same here.
 

You slaging off ppu and physics then you post a link to that haha, i not got a ppu but i would like physics to take off in games.
You happy to download demo's knowing half of them will never apear for years if ever and never in a game and you say it doesnot matter it not a game,you play with your demo's i want physic that work in a game.
It the same as the nvidia head that fine if all you want is a head on the screen and nothing else but put it into a game and see how it runs it will be ages before it in a game at decent frame rate and ati no better really.
I am getting sick and tired with all the demos about which more than half never see the light of day at least with ppu they are trying to go forward.
Yes it cost to much at moment for me but i not going to **** someone else off for buying one.
At the end of it cellfactor is out and like pottesy said show us a game that does physic aswell with out a ppu.
http://forums.facepunchstudios.com/showthread.php?t=309541
 
There is the same reaction for ppu and physic as for voodoo cards when they came out,people moaning about buying a extra videocard for 3d only and look at it now.
 
Last edited:
3dfx Voodoo vs Matrox Mystique wasn't it? That's casting my mind back... but yes pretty much.

yes it is i had both mystique became my 2d card and voodoo was my 3d card

I hope unreal 3 does phsics well and not just a bolt on
 
Well first of all that’s fake cloth that doesn’t tear or rip its only semi-dynamic and is also suffers from massive slow down which proves my point its not do able on the CPU at useable speeds. They said “This is completely useless in multiplayer because it would lag” they also said “'cloth' just bumps into itself, making it slow down (looks pretty weird when falling)” and The physics beeing buggy, can't really be fixed, its because havok's physics engine isn't build to make a **** load of constrained entitys, things you shouldn't do are pulling the cloth apart and dropping it from large height.”

So not only is it super slow it’s buggy and not fully dynamic. You can also not tear or rip it.

Watch the video as well from 18 seconds in the cloth is falling though objects and not acting towards it surroundings. It looks and acts terrible. It doesn’t really look like cloth and its not falling on and around objects very well. It also has other problems like “All the balls on the cloth would probably have the same weight so they would all fall down at the same speed. Some parachute.” Basically the cloth doesn’t act like cloth.





“It doesn't matter if Gmod isn't a game. Its using the Source engine so we know it can be done.”
It does matter. I never said it couldn’t be done full stop, I said it couldn’t be done at useable speeds in a game. Which you’re proving if you read about it, they said it’s not useable at useful speeds and buggy which cannot be fixed. Sure you can make cloth if you put 95%+ of your systems resource’s doing cloth and just cloth but you cannot do anything else so it cannot be used in a game. Same for the Face demo from Nvidia you don’t get detail like that in a game as it would just kill your computer.
http://www.istheshit.eu/meshcloth/images/MeshCloth_Beta1_5.jpg
http://www.istheshit.eu/meshcloth/images/MeshCloth_Beta1_6.jpg
 
No game out currently makes use of proper fluids and cloth simulation. Fluid dynamics and cloth solvers require a lot of resource and theres no game which looks like it is properly making use of how cloth or fluids would be simulated in an industry standard 3d content creation package. Take the application Real Flow for example, http://www.nextlimit.com/realflow/ (this may interest you Pottsey), these types of complex solver are not present in games. The cloth I've seen in games is merely simulated by soft body dynamics, the game I first saw this being used in is the original splinter cell and it looked great. Years on, nothing has really changed. The cloth is cell factor is not being properly solved, it is a soft body controlled by a few scripts. A soft body is a polygon object, usually a 2d plane, with weighted particles locked to each vertex. Real cloth effects are simulated by a series of springs or a more modern approach, rigid links and then softening them if any stretching is desired. All a cloth object is in 2d space is a simple rope. When I see games where characters are wearing full clothing that ripple and flow 100% correctly as they move like I've seen in Maya for years, then ill say its "proper cloth". Check out Star Wars episode 2, where Yoda is battling Count Dooku and his cloak is swirling and moving realisticly. This is what I'm talking about. When I see fluids being simulated the same way as done in Real Flow but in real time, then I'll say games have proper fluids.
 
Last edited:
Nice job there queamin in making 4 posts in a row in the space of 13mins. You realise there is an edit button? And what are you, Pottsey's bum-chum?

Fair enough Gmod isn't the greatest example. But how can you laugh when one person created that cloth tool. Ageia have hundreds of people under their belt working for cloth and liquid physics and of course they look good. I never said they didn't. But all im saying is that it is possible to simulate such things without a ppu.
 
Pottsey, was wondering if you could do me a favor. Could you benchmark cellfactor with your PC, iv seen some online but not a 'home system' as it were.

My ponderment is from, ok CPU does some things, GPU does graphics, PPU physics, but just HOW FAST is the PPU, i.e. on a theoretical 'core 10' with Nvidia 12000 running cell factor without PPU at 200fps, what would happen if PPU was enabled, ~40fps all time? hows this helping? yes I KNOW it has more to deal with.......but if you could 'force' it to run all physics enabled on a 'core10' cpu, it would run faster.........

Which leads me to my next ponderment, will ageia release faster and faster versions of its PPU ? wouldn't that make a PPU 'redundant' ? Why would people buy a PPU card to play the latest games? may as well just buy a console......of course the same could be applied for cpu and gpu, but they do have more general uses, I just dont think the public would go for cpu , gpu and then something else as well (PPU) to play a game

I think it will ONLY work if the CURRENT PPU is more or less the ONLY one they will ever release (Id definently buy one if I know it will be the same one they will still sell in 5 years)..........and when its price drops it should be adopted more, and will actually be used, a 3rd 'add in' card thats more or less 'non changing' would be a great angle to take.........also programmers would be able to eek every last drop of performance out of it........which if everyone has the same card in there pc should make for some useful and (at long last) well written game code

Just my 2p and how I think it should go
 
Not to sure if Cellfactor can be benchmarked in the home version. Review sites use the press version with more options. Anyone know the command to start a benchmark? The other problem is you cannot force the extra physics on like you can in the press version or run the high physics maps with the CPU. So when I turn off the PPU I lose physics making an unfair comparison.

I could use the olf Cellfactor benchmark if that’s ok or did you only want the new game?
 
Last edited:
Why dont you just record the fps using fraps benchmark on one of the levels that can run w/o a ppu? like the one on the rig looking thing :s with all the boxes in the middle...would be a good benchmark to do that level for 2 mins with fraps recording, once with ppu and once w/o it
 
I can give it a go but it wouldn’t be accurate. It would mean turning the bots off, I wouldn’t walk precisely the same path. It would also be benchmarked in the low physics levels when I am sure it’s the high physics levels everyone’s interesting it. Going try and get hold of the press copy first.

As I understand it without the PPU the physics are lowered as well unless you have the press copy. I don’t like the idea of running a benchmark with a PPU on max settings and without a PPU with lower settings.
 
tbh my biggest issue with PPU's is it's just another thing we have to spend out for on top of the CPU and GPU. I for one will not buy one, not from AGEIA or by grabbing an additional GPU for physics. It should be incorporated into the GPU or CPU and touted as technological progress.
 
WARNING: I HAVE A GOING-OFF-ON-A-TANGENT PROBLEM

I thought they where shooting blocks of blue snow down the hill in that gary's mod, and the water..... well it's a good effort. Of course physics can be done by the cpu or gpu, and it has been doing in games for years (on cpu anyway), but to reinterate on a previous post, only on basic things such as; 1 or 2 boxes that u need to move.

I know the PPU is basically only in embryo form right now, and it is pretty much a sort of break through in the gaming sense. I realise the water looks like small spheres, but it is all interactable, but not pretty as i think making the gpu render lighting and textures on all that lot will actually kill it. But the PPU can also be used in the study of science, like calculations in the theoretical crumple zones in a car at some point. It can probably even calculate aerodynamics for military use, counting in air particals as part of its physics.

There is an extremely big picture here regarding the future of the PPU, sure ageia 'might' retract this project from the gaming scene, but after awhile you'll definitely see it back under a different company willing to pick up where it left off. This seriously is the future, and there's no denying that, along with AI and this excites me. Fair enough you can discard something that will never catch on, like a solar powered torch, but the PPU is the next logical step towards a more complete gaming experience we all crave for.

If more flying and colliding boxes is all you see, then isn't that a bit of a short sighted view? Future development could see fully realistic destructable environments, including the floor.

Theoretically speaking if you had an artillery implacement in the forest and you were bombing a building 1k away which you can see with your binnoculars. Bearing in mind the wind resistance, you set your target up and fire, instantly all the dust around you would rise. You manage to hit your target and watching as it implodes, and all the debris being initially sucked in by the vacuum and then being instantly dispersed as the building crumbles to the ground, and having the dust being carried off in the direction of the wind. Bits of building (and bits of people) settling on nearby trees etc. This is just a basic example but doesn't this sort of thing float your boat?
 
Jihad said:
I just don't see the big deal with some cloth that gets shot into pieces, I can't recall every game I've seen with destructible cloth but there's been some atleast.


red faction 2 has better physics that cell factor and that doesn;t even need a 1ghz cpu to do it.

agiea is crap and always will be. nothing they have ever done will be of any good since all they are after is cash. thier physcis implementation is utter crap, it kills performance and its just a waste of time. and to top it off the physics in cellfactor are just utter poo, all objects behave like shoot 5 footballs. :rolleyes:
 
Cyber-Mav said:
red faction 2 has better physics that cell factor and that doesn;t even need a 1ghz cpu to do it.

agiea is crap and always will be. nothing they have ever done will be of any good since all they are after is cash. thier physcis implementation is utter crap, it kills performance and its just a waste of time. and to top it off the physics in cellfactor are just utter poo, all objects behave like shoot 5 footballs. :rolleyes:

Sorry, but red faction 2 had crap physics. It supposedly had a 'revolutionary' feature that allowed for loads of destructable scenary (walls ect.). But that just turned out to be pre-determined walls that could be shot through/ blown up (while looking naff at the same time). Other than that (which don't really count as physics) there was nothing special about the physics on RF2. I don't think the physx cards are any good, but I do think some of the stuff on show in cellfactor is mighty impressive. Waaay ahead of RF2.....
 
Cyber-Mav said:
red faction 2 has better physics that cell factor and that doesn;t even need a 1ghz cpu to do it.

agiea is crap and always will be. nothing they have ever done will be of any good since all they are after is cash. thier physcis implementation is utter crap, it kills performance and its just a waste of time. and to top it off the physics in cellfactor are just utter poo, all objects behave like shoot 5 footballs. :rolleyes:

LOL it's a game, you hardly expect someone who has psychic powers to be able to control them with a limited control system such as a mouse and keyboard.

Have you played it yet? If you shoot a wall at any angle, it will deflect and blow apart in an ALMOST accurate way........


Oh also about the benchmark for CellFactor on a home system, if you go into graphics there is an option which gives you a whole read out of the details. Record it using wme or similiar and then post it - I can't as I've recently ran into problems with 2 of my hard drives and have a total of about 100mb free on my raptor :(
And Havok isn't exactly realistic either, just look at TDU, cars and things bounce all over the place, just as the boxes do in HL2 with the grav gun.
 
“thier physcis implementation is utter crap, it kills performance and its just a waste of time.”
Why do you keep spreading lies like that? I might not have shown benchmarks of GRAW since the performance bug was fixed but I did show the other games like COH and the cellfactor benchmark.

You have yet to show one recent benchmark where the PPU kills performance. The only benchmarks where it kills performance are 8+months old before the performance bug was fixed. Eveything since then points to a performance boost.

EDIT:Found some new GRAW benchmarks with the new drivers.
hrrm filters blocked out my links. I also read the date wrong not as new as I thought. Serching for more that are not blocked

http://www.pcper.com/article.php?aid=403&type=expert
Cellfactor new.

EDIT2: I give up trying to find new GRAW benchmarks. I found some older ones that don’t kill performance or boost performance either. But no new benchmarks with the new faster drivers. Eveythings a year old benchmark wise.
 
Last edited:
Hmmm.

I dunno, its not crap if you play the few games that use it but it is a waste of money at this stage, which is what i was arguing earlier.

There is NO point buying them now, at all, untill more games support the ppu. its over-priced imo too, if they want more people to buy them, they really need to nose-dive with the prices

I just dont think the "real" cloth and "real looking" water is worth the price tag at the moment. And having played the cellfactor game a bit more, i dont see anything special on the two available levels that inspire me to part with my money.

same with GRAW demo (although i hear this is badly coded or something?)

Edit:

Cyber, post some benchmarks proving the ppu SLOWS down operations, please

and lay off Pottsey, he really knows his stuff and has outsmarted me a few times, although, the fact theres no RECENT benchmarks of this ppu does strike suspicion for me
 
Last edited:
Back
Top Bottom