• 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.

PhysX Accelerators

skullman said:
Software engine or not, it specifically says on the CryEngine 2 website "without the need of specialized coprocessing hardware" - which is what matters. It also says that the physics engine will benefit from multicore cpu's.
Click to expand...

Imagine a massively interactive environment versus a moderately interactive one, for say, £150 every two years for a physics card upgrade. Would you pay that? Surely nobody is arguing that a CPU can perform these calculations better than a PPU? I think Pottsey has argued very well that this is not the case.
 
Im not sure if its been said, but i have seen benchmarks on this card vs no ppu, and the difference was amazing...there was barely any at all. There is a way to speed the games that use it up without actually having the ppu in your computer at all, been looking for it again for 2 days now and cant find the damn thing...will post it if i come across it again though.

Anyways, i dont see the point in slashing out £150 for a card that makes things move a little bit better...think about it, it makes cloth look and move more realistically, but are you gunna notice when in a firefight against a few enemies? Not really, thats the main reason why i wouldnt but one. Same with water really.

Havok does the job, and imo does it well.
 
“Im not sure if its been said, but i have seen benchmarks on this card vs no ppu, and the difference was amazing...there was barely any at all.”
How can you call a 30% fps boost barely any at all? Or like most people are you ignoring recent benchmarks and games and only looking at the year old benchmarks with old buggy drivers?

I have a challenge to everyone who says the PPU is slow or does not boost FPS. Find benchmarks in new PPU games with new drivers that show a slowdown. 6+ month Old drivers and old patch's of games dont count.

EDIT: CellFactor: Revolution Ships May 8.
 
Last edited:
Pottsey said:
“Im not sure if its been said, but i have seen benchmarks on this card vs no ppu, and the difference was amazing...there was barely any at all.”
How can you call a 30% fps boost barely any at all? Or like most people are you ignoring recent benchmarks and games and only looking at the year old benchmarks with old buggy drivers?

I have a challenge to everyone who says the PPU is slow or does not boost FPS. Find benchmarks in new PPU games with new drivers that show a slowdown. 6+ month Old drivers and old patch's of games dont count.

EDIT: CellFactor: Revolution Ships May 8.
Click to expand...
It was 5fps difference, which is barely noticeable anyway.

It was someone who figured a way to get the performance of the card without having the card at all, as i said, i cant find it again, but i will keep looking.
 
Greenboi said:
It was 5fps difference, which is barely noticeable anyway.
Click to expand...

Which benchmark are you referring to?

Greenboi said:
It was someone who figured a way to get the performance of the card without having the card at all, as i said, i cant find it again, but i will keep looking.
Click to expand...

Are you saying that someone worked out how to get the same performance from a CPU as you get from a PPU in all cases or in just one game?

If you are saying that they found a way in all cases then I seriously doubt that they did. As has been mentioned on several occasions you will not get the same performance from a CPU for physics calculations compared to dedicated hardware (PPU or GPU).
 
I have a link, but its to another forum (and its actually not the benchies i was originally referring to)

So, i assume i cant post the link to the forum, but it shows that the biggest difference with and without the PPU is

19fps with
12fps without

Also, forgot to mention, that in a few screenshots the actual FPS was LESS, yes, LESS with the ppu than without, tests were done on Cell Factor

Worth £150? Hell no, id rather put that to my current GPU fund and get me a R600 high end card..

edit: One bench: http://www.hothardware.com/viewarticle.aspx

On a game that people have actually heard of, unlike Cell Factor..
 
Last edited:
Greenboi said:
I have a link, but its to another forum (and its actually not the benchies i was originally referring to)

So, i assume i cant post the link to the forum, but it shows that the biggest difference with and without the PPU is

19fps with
12fps without

Also, forgot to mention, that in a few screenshots the actual FPS was LESS, yes, LESS with the ppu than without, tests were done on Cell Factor

Worth £150? Hell no, id rather put that to my current GPU fund and get me a R600 high end card..
Click to expand...

Did these benchmarks have exactly the same effects running when the PPU was enabled and disabled. What PPU drivers were being used?
 
“edit: One bench: http://www.hothardware.com/viewarticle.aspx
On a game that people have actually heard of, unlike Cell Factor..“
Those Ghost Recon benchmarks are 1 year old. Not only that but they used the old drivers which had performance problems and the game was before the patch which fixed various other performance problems.

Considering the game has known problems that have since been fixed and the drivers are no longer slow its not fair to judge the card how it used to be 1 year ago.

Please stop judging the PPU on the first 2 sets of drivers. The new games and drivers give up to and sometimes more then a 30% fps boost with extra effects.

EDIT: Checking the benchmarks again they didn’t even use final drivers. Those are benchmark with beta drivers. Doesnt seem fair to me to use 1 year+ old benchmark with beta drivers.





“Did these benchmarks have exactly the same effects running when the PPU was enabled and disabled. What PPU drivers were being used?“
No the CPU had less effects. That another thing that bugs me about reviews if your going run the CPU with less effects make it clear. To many people look at the benchmarks and think the CPU is running the same effects as the PPU. Its like running benchmark on an ATI card with 0 FSAA and an Nvidia card with 4x FSAA not saying anything about x4 FSAA then going look the ATI card is as fast or faster.
 
Last edited:
If less effects are used when running CPU only, then how are we supposed to know how well a CPU could actually handle such extra effects or not, or how much benefit Ageia's hardware actually offers?
 
Pottsey said:
“edit: One bench: http://www.hothardware.com/viewarticle.aspx
On a game that people have actually heard of, unlike Cell Factor..“
Those Ghost Recon benchmarks are 1 year old. Not only that but they used the old drivers which had performance problems and the game was before the patch which fixed various other performance problems.

Considering the game has known problems that have since been fixed and the drivers are no longer slow its not fair to judge the card how it used to be 1 year ago.

Please stop judging the PPU on the first 2 sets of drivers. The new games and drivers give up to and sometimes more then a 30% fps boost with extra effects.

EDIT: Checking the benchmarks again they didn’t even use final drivers. Those are benchmark with beta drivers. Doesnt seem fair to me to use 1 year+ old benchmark with beta drivers.





“Did these benchmarks have exactly the same effects running when the PPU was enabled and disabled. What PPU drivers were being used?“
No the CPU had less effects. That another thing that bugs me about reviews if your going run the CPU with less effects make it clear. To many people look at the benchmarks and think the CPU is running the same effects as the PPU. Its like running benchmark on an ATI card with 0 FSAA and an Nvidia card with 4x FSAA not saying anything about x4 FSAA then going look the ATI card is as fast or faster.
Click to expand...

I quite agree.

Everytime a thread about Physx comes up people always harp back to what the performance was like when it was first released and was benchmarked with GRAW. They completley ignore anything that has happened since and in a lot of cases entirely miss the point of what the PPU was designed to do.

You also here lots of arguments in favour of the proposed solution from ATI and Nvidia of running physics simulations on the GPU. They ignore the fact that Havok FX has been out for ages but has not, as far as I know, been supported in a single game yet. I believe the first will be Hellgate London.

The latest argument is that high level physics can be achieved by running on a quad core CPU and people point to Alan Wake as an example of this. Whilst using a quad core CPU will definitley increase the complexity of the physics you can use they are not anywhere near as complex as those that can be achieved in real time by using dedicated hardware be it a PPU or GPU.
 
A random name said:
Pottsey, you really need to just stop trying to defend your purchase so much :\
Click to expand...

no, you need to get off his back unless you have a valid arguement. It's cool to hate physx right now as is obvious by this post you've just made.


everything he has said is correct and i challenge you to counter his points with relevant, up to date information. otherwise just keep quiet on the subject - its getting BORING.
 
Out of curiousity - is there a list of released games that make use of the physx card. Also - can someone answer me this - is the fan on the card temp controlled or does it run full whack all the time and is it loud ? thanks.
 
Flanno said:
Out of curiousity - is there a list of released games that make use of the physx card. Also - can someone answer me this - is the fan on the card temp controlled or does it run full whack all the time and is it loud ? thanks.
Click to expand...

As far as I know its whiney and not speed controlled.
 
Just looked at the ageia website and ghost recon2 will use the physx card, as will unreal tournament 3. a few more games of that calibre and I would definitely consider buying one.

Wondering how easy they are to modify with something like a zalman vf900.
 
I can't believe that this is still going. I remember having this conversation a year ago.

Last June pottsey assured everyone moaning that there weren't any games that use PhysX that:
pottsey said:
But I do agree there are not enough games out luckily that will change with time as a lot are due out this year.
Click to expand...
What happened?

fini
 
You must log in or register to reply here.
Back
Top Bottom