Your right, I got it and the 8600GT mixed up with the GTS. And the 8600GTS was on average about as good as a 7900GS, still a fail in my book as the GTX and both GTs(7900/7950) were faster. And all 3 are conciderably faster than a 7900GS.
AMD: "[Developers use PhysX only] because they’re paid to do it"
- Thread starter Richard
- Start date
Your right, I got it and the 8600GT mixed up with the GTS. And the 8600GTS was on average about as good as a 7900GS, still a fail in my book as the GTX and both GTs(7900/7950) were faster. And all 3 are conciderably faster than a 7900GS.
@ kaotik that was me
Nvidia's Rod Taylor
"Our argument to those guys that are developing their own engines or using others is that we understand why they've developed their own physics engine, as they may have created something unique or which was wasn't possible with earlier versions of PhysX. We understand that.
However, we have two responses. The first is that if they want us to, we'll help them run their physics engine on the GPU because we can do more complex parallel calculations faster than you can on the CPU. Secondly, Nvidia is taking PhysX very seriously, we have hundreds of people working on it and we are investing heavily in it so developers can be sure that its going to get better and better at a faster rate than many can expect to develop physics independently."
Nvidia's Rod Taylor
"Our argument to those guys that are developing their own engines or using others is that we understand why they've developed their own physics engine, as they may have created something unique or which was wasn't possible with earlier versions of PhysX. We understand that.
However, we have two responses. The first is that if they want us to, we'll help them run their physics engine on the GPU because we can do more complex parallel calculations faster than you can on the CPU. Secondly, Nvidia is taking PhysX very seriously, we have hundreds of people working on it and we are investing heavily in it so developers can be sure that its going to get better and better at a faster rate than many can expect to develop physics independently."
As willard asked these posts be made here
How about giving us a link to a Havok based preview where havok does something DIFFERENT than the same thing over and over again. The Eurphoria tho was very cool tech. Havok is only good if your a piss poor dev who can't be bothered to take the time to do reactionary physics. I'm sorry, but if PhysX can do it, if CryTek can design from ground up, not once but twice now, in house reactionary physics, why the hell is it every single dev who uses havok can't do the same thing? It can only be 1 of 2 reasons for, lazy devs or massive limitations of Havok to do reactionary, more life like based physics.
How about giving us a link to a Havok based preview where havok does something DIFFERENT than the same thing over and over again. The Eurphoria tho was very cool tech. Havok is only good if your a piss poor dev who can't be bothered to take the time to do reactionary physics. I'm sorry, but if PhysX can do it, if CryTek can design from ground up, not once but twice now, in house reactionary physics, why the hell is it every single dev who uses havok can't do the same thing? It can only be 1 of 2 reasons for, lazy devs or massive limitations of Havok to do reactionary, more life like based physics.
Hmh? Last time I played Havok-accelerated game, it didn't react same way everytime, like you seem to suggest.
And I think Dave was referring more to DMM, which is being integrated with Bullet for "Open Physics initiative"
Yes, indeed. Odd post from XMAN considering there is a lot of elements demonstrated in games for the first time in there and wasn't even related to the Havok element of the title.
Nope, rigid bodies physics that dissapear before 10 seconds is never seen before in CPU physics...oh wait..
Force Unleashed is not something I would cheer, hell even Cellfactor did it better...in 2006.
Nope, rigid bodies physics that dissapear before 10 seconds is never seen before in CPU physics...oh wait..
Force Unleashed is not something I would cheer, hell even Cellfactor did it better...in 2006.
Force Unleashed is a console game. It did alot of things no console game did before it. In fact its a much better show of physics than any physx console game.
That is not saying very much...bad physcis is bad physics.
But you should look op some Brothers In Arms: Hell's Highway console videos...I saw nothing in Force Unleashed that surpassed that *shrugs*
And please define what you mean by physcis?
Rigid bodies?
Effects physics?
Ragdolls?
Panajev2001a
Veteran
Personally the best recent Physics tour-de-force I have played is Red Faction: Guerrilla. Was the physics engine for deformation, demolition, etc... all developed in-house part of the GeoMod project? Having to actually model structures carefully so that they do not actually fall down due to the physics engine being always active (the game looks very good too) must be something new for designers...
All I see is the same CPU physics as always, simple rigid boides that dissaper in time...with a lot of clipping isusses?
<cough>
(metacritics score based)
And, I've seen that. The more interesting graph is below that showing a slow and steading decline of Havok basd games being released while PhysX on the rise.
Must not have played BC2. Building fall apart the same reguardless of what you use to destroy it. NOT very realistic and the animation is the same no matter the angle from which you do it. Lazy Dev coding or Havok limitation, either way, not very good.
And Dave, where oh where is ATIs GPU based physics acceleration answer you were working on? 3 years later and nothing really to show for it.
BF BC 2 does that because it needs to be played over a range of systems.
You can't go crazy with physics in heavy multiplayer games . You might have someone on a 2ghz dual core cpu and some one on a 6 core cpu at 4ghz .
understandable, but I really dont think it would have added much more complexity to it to add a bit more variance to it so that it doesn't look the same when you blast a wall down from a 45 degree angle as it does head on.
IT would have to update all clients at the same time. For peer to peer gaming i don't see how its possible without massive slow downs for gamers.
People could move on from 56K...i have a 15/15 Mbit Gbond SHDSL line(look into SHDSL lantency
), should be more than ample to update physcis over WAN...hell even Cellfactor didn't suffer on 10Mbit LAN...People could move on from 56K...i have a 15/15 Mbit Gbond SHDSL line(look into SHDSL lantency
People have. You'd still have the problem of one users computer having to update all the phsyics. Or the server having to update all the physics. But i doubt anyone is going to want to foot the bill for such a powerfull system.
It could be done in 2006 in Cellfactor...and you didnd't need the best of the best hardware, just a PPU....so you are sayng it would be more costly today?
Parsing x,y,z cordinates is NOT computional heavy, it depends on...*TA-DAA*...bandwith....your argument is illogical