• 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.

2K Games, EA hop on the PhysX bandwagon

LOL yeahhh.
My bad! I did not realise that 99% of all game code has all been psychics code. :rolleyes: :rolleyes:

It's not just about physics, CUDA/Physx technology is making CPU's more and more redundant because GPU's are simply better at heavy number crunching, take a look at Folding@Home performance etc.

AMD know this so it's not in their interest to promote CUDA/Physx, infact quite the opposite, they want to keep CPU's as the main workhorse for their own selfish gain, even though NVidia are offering us (the consumer) a far superior physics technology than anything a CPU can offer.
 
Last edited:
LOL yeahhh.
My bad! I did not realise that 99% of all game code has all been psychics code. :rolleyes: :rolleyes:

"physics" code makes up a reasonable proportion of game code - probably around 20% you can't really put hard figures on it.

In most games "physics" is little more than a couple of simplified movement types (linear and mavity), rotating the objects heading (vector) and doing simple collision detection without orientating the object to the collision.

Things have moved to the point where this is no longer adequate and so you have software engines like havok, physx, tokamak, newton and so on - but most of these are limited to around 200-300 primitive (box, sphere and cylinder) bodies and no complex physics such as concave interaction, advanced convex to convex/concave, fluid dynamics, cloth, advanced field effects, limited dynamic fracturing and so on... game developers do want to use these features and newer games are going to require them.

And thats not even taking into account joints/constraints, etc.
 
It's not just about physics, CUDA/Physx technology is making CPU's more and more redundant because GPU's are simply better at heavy number crunching, take a look at Folding@Home performance etc.

AMD know this so it's not in their interest to promote CUDA/Physx, infact quite the opposite, they want to keep CPU's as the main workhorse.

One client & you think the GPU can crunch anything.
Everything has it limitations.
The CPU Is the main workhorse & will remain so for the foreseeable future.
 
It's not just about physics, CUDA/Physx technology is making CPU's more and more redundant because GPU's are simply better at heavy number crunching, take a look at Folding@Home performance etc.

AMD know this so it's not in their interest to promote CUDA/Physx, infact quite the opposite, they want to keep CPU's as the main workhorse for their own selfish gain.

AMD's been backing OpenCL and DirectX 11's compute language (vendor independant GPGPU solutions) more than their own Stream SDK. Unlike Nvidia who believes their solution is the be-all and end-all, well, end-all good prices if they truly get it off the ground above the middle-ground competitors.
 
"physics" code makes up a reasonable proportion of game code - probably around 20% you can't really put hard figures on it.

In most games "physics" is little more than a couple of simplified movement types (linear and mavity), rotating the objects heading (vector) and doing simple collision detection without orientating the object to the collision.

Things have moved to the point where this is no longer adequate and so you have software engines like havok, physx, tokamak, newton and so on - but most of these are limited to around 200-300 primitive (box, sphere and cylinder) bodies and no complex physics such as concave interaction, advanced convex to convex/concave, fluid dynamics, cloth, advanced field effects, limited dynamic fracturing and so on... game developers do want to use these features and newer games are going to require them.

And thats not even taking into account joints/constraints, etc.

20% is not a significant portion.

In most games "physics" is little more than a couple of simplified movement types (linear and mavity), rotating the objects heading (vector) and doing simple collision detection without orientating the object to the collision. & people love it & don't care how it was done
 
Last edited:
20% is not a significant portion.

In most games "physics" is little more than a couple of simplified movement types (linear and mavity), rotating the objects heading (vector) and doing simple collision detection without orientating the object to the collision. & people love it & don't care how it was done

Unfortunatly for "people" its not capable of powering future games - sure most people don't care how its done - but thats why most people in this thread should shutup tbh because they don't appreciate the application of hardware physics in the game development world and how limited the old methods are.
 
Unfortunatly for "people" its not capable of powering future games - sure most people don't care how its done - but thats why most people in this thread should shutup tbh because they don't appreciate the application of hardware physics in the game development world and how limited the old methods are.

You can not demand people to appreciate something before it is made.
I was pro PhysX when it first arrived as i could see its potential. but i have yet to see to this very day anything in any game thus far making any worth while use of it & im frankly tied of waiting & really don't care any more.
Until there are mainstream games that makes good use of it as far as im concerned PhysX may as well not exitst.

I don't remember you pushing for PhysX on OcUK when PhysX first came out & i only see a few NV owners pushing for it & no PPU card owners who probably have lost hope & unlike you the PPU card they bought does not happen to have gfx so is of little use.
 
Last edited:
In most games "physics" is little more than a couple of simplified movement types (linear and mavity), rotating the objects heading (vector) and doing simple collision detection without orientating the object to the collision. & people love it & don't care how it was done

...and what was "graphics" like before we had hardware acceleration for that?

Here's a reminder:

wolfenstein.gif


Good old CPU acceleration. ;)
 
I don't think anyone really has a problem with hardware accelerated physics, the concept makes sense, sure, but who really wants to have it give a huge upper hand to one vendor because they were all, 'IN BEFORE YOU GUYS'?
 
...and what was "graphics" like before we had hardware acceleration for that?

Here's a reminder:

wolfenstein.gif


Good old CPU acceleration. ;)

A poor comparison for you argument as it had a much more immediate & dramatic effect in the enthusiast market unlike PhysX.

MY argument is not that PhysX could not have a some what similar effect but by its very nature good PhysX is harder for the avg consumer to notice than good gfx.

And just because dedicated gfx acceleration took off so well does not automatically mean PhysX will.

And last but not least the Move to dedicated gfx acceleration did not lower the demand for the CPU.
 
I don't remember you pushing for PhysX on OcUK when PhysX first came out & i only see a few NV owners pushing for it & no PPU card owners who probably have lost hope & unlike you the PPU card they bought does not happen to have gfx so is of little use.

At the time software physics engines could handle all the workload that was really feasible in game engines at the time and it just wasn't worth it for the odd incidental effect - now we are at a stage where we can use incidental effects throughout the game world and more involved effects and games are moving to the stage where they need more physics objects and techniques than can be sensibly handled on the CPU in realtime... and so in some cases (and its only going to become more increasing) developers are marking time or cutting great features out of games (that are feasible now on current generation hardware acceleration) because it would mean a very narrow market...

I'm not pro-physx specifically anyhow - I'm very pro hardware physics acceleration and very anti companies that not only refuse to work with nvidia to develop this, not only drag their own heels over getting something working... but also actively seek the demise of this solution... they are doing no one a favor.
 
At the time software physics engines could handle all the workload that was really feasible in game engines at the time and it just wasn't worth it for the odd incidental effect - now we are at a stage where we can use incidental effects throughout the game world and more involved effects and games are moving to the stage where they need more physics objects and techniques than can be sensibly handled on the CPU in realtime... and so in some cases (and its only going to become more increasing) developers are marking time or cutting great features out of games (that are feasible now on current generation hardware acceleration) because it would mean a very narrow market...

But your not developing any of those games so you will have no idea about what had to be left out & that games next year will not be any good because of no accel physx.
accel physx, flashy GFX does not make the game.

I'm not pro-physx specifically anyhow - I'm very pro hardware physics acceleration and very anti companies that not only refuse to work with nvidia to develop this, not only drag their own heels over getting something working... but also actively seek the demise of this solution... they are doing no one a favor.

Seeing as you don't work for or have shares invested or have inside info of the other parties & no worry of direct financial loss you have no right to ask this of them nore do you know the facts for them not working with NV.
 
your making a lot of assumptions about who I may or may not work for :D

Your making allot of assumptions about the reasons for other parties not working with NV & if your involved in anyway then your not impartial & can have any number of alteriamotives for pushing physx that only have your best interests at heart & would require more than just your world of proof that anything your saying is nothing more than marketing spin.
 
I have to generalise on why companies don't get themselves involved with physx and other hardware solutions as theres a huge number of variables and I only know in depth for a limited number of companies.

And I'm not going to start going into proof of who I might or might not have a working relationship with but I will state that I am involved with the game development industry at a number of levels and while I'm not entirely impartial I do feel qualified to comment on things from the game development perspective.
 
I have to generalise on why companies don't get themselves involved with physx and other hardware solutions as theres a huge number of variables and I only know in depth for a limited number of companies.

And I'm not going to start going into proof of who I might or might not have a working relationship with but I will state that I am involved with the game development industry at a number of levels and while I'm not entirely impartial I do feel qualified to comment on things from the game development perspective.

When you come in to a General computer forum & not a Developer forum & not say anything of what your involved in which is your right, people can only assume from what you type because there is no other basis.
Tho i don't your claims.

And seeing as to your possible involvement which would mean it would be impossible for you to be seen as impartial, my discussion on this matter with you is over.
 
if you wanna talk about beta testing (and this isn't where I would draw my "credentials" from) - I've been involved in private (not public) testing of around 1 in 5 of the titles released over the last 5-6 years or so including city of heros, enemy territory quake wars, hellgate london and a good many others. I've also done contracted level design for a number of smaller titles and been involved in the programming side of more than one title.
 
Last edited:
Back
Top Bottom