• 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.

PPU more then 100% faster then CPU in UT3 benchmark here.

sorry pottsey, but it hasn't change an iota. firstly, claiming 30fps when its not, 26fps would need close to a 20% speed boost to be counted as 30fps. 20% is not a small difference.

the WHOLE point you keep making is that PPU negates the need for cpu power, so why would an ultra fast dual core make a difference. ut3 DOES NOT ever use high cpu usage on my quad core at high settings, ever. its not a cpu intensive games, none are.

what i want to know is, what effects does the ppu give in normal maps, over a non ppu setup? is it the same completely and utterly pointless 7 extra particles when crap explodes? effects wise it does nothing. their whole point is to make maps YOU CAN'T PLAY, by making it SO intensive on the ppu that its ridiculous on a normal cpu, and still crap on a ppu. the whole point you miss here, is with a quad core, the ONLY difference will be on the PPU specific maps ppu speed will stay the same, but quad core speed should increase the dual core speed massively, getting much closer to the ppu, only further pointing out how freaking useless ppu's are.

you argue they didn't use a high end system, i contend that they didn't use a quad core as it would have only narrowed the performance advantage in the ppu maps. they did it on purpose, i wonder what the difference would be between quad core and ppu on the two maps, probably much closer, dual core + £88 would get you a quad core, which would be more useful in infinately more situations than a ppu.

again, yet again, PPU FAILS.

oh and another PPU game comes out, Cellfactor was claimed to have been amazing right, a great game, which became a couple levels tech demo that basically no one spent time on.

their second game, its a 437.2mb download, are you really thinking its going to be such a break through game that everyone will get a PPU? or yet again, for the millionth time show that adding an extra physics system thats completely unnecessary into a game takes more time than its worth, and ppu's will only ever be used to add pointless effects that don't add realism at all(oh look that barrel exploded in 12 pieces rather than 8, my game has completely changed, because the game is called "barrel explosion, count the pieces to win"), or used to make maps that aren't part of the normal game, run marginally better on ppu's than DUAL CORE cpu's, but maps no one will play more than a couple of times. or a couple of games/tech demo's that will be largely ignored because they get massively cut down as ageia's own people can't produce anything in a realistic time frame because of the overcomplexity of their system.
 
Last edited:
Whilst this is a good performance difference 26.1fps in UT3 is still not playable (given thats an average)
Yup, that's awful performance for an 8800GT and Core 2 Duo.

Funny how Pottsey says the low framerates in Crysis are due to physics and not the GPU, yet the lower dramerates in the PPU maps are coz of the GPU/CPU. Please make your mind up Pottsey.
Well said.
 
“Exqueze me? Not high end hardware?“
Quad core = high end
Dual core = mid end.

HAHAHAHAHAHAH, a C2D @ 3.4ghz aint mid end by any stretch of the imagination! At stock maybe, AMD64 939/Old AM2 is mid end now.

Unless I got the GPU’s muddled up its
8800 GT = mid end
GTS/GTX = high end.

8800GT when I last looked was quite close to its "better" sisters, high end too me thinks.




“£88, for still under 30fps.“
Its more like quad core = under 15fps
Quad core + PPU = over 30dfps.

Normal map without either = 60

Hmm yes I know what I want!!! 60 FPS Vs 30FPS with some slightly better physics hmm tough!


Sorry but its still a load of ****.

I'd buy one when/IF Microsoft makes proper use of it in FS-X to calculate crash physics/weather/traffic patterns...

But oh no, have they made the CPU do it all? YES!!

Ok I'll keep my £88
 
“oh and another PPU game comes out, Cellfactor was claimed to have been amazing right, a great game, which became a couple levels tech demo that basically no one spent time on.”
How many times do I have to correct you on that? It’s not a couple level tech demo.
It’s got more then a couple of levels and it’s a full game. It has way to much to be a tech demo.


“again, yet again, PPU FAILS.”
The PPU boosts FPS on all maps over the CPU that’s not PPU fails.




“i contend that they didn't use a quad core as it would have only narrowed the performance advantage in the ppu maps. they did it on purpose,”
Clearly you didn’t read what the was written. Even with the PPU the game is CPU limited and quad core will improve FPS with a PPU. Also the CPU usage shows the dual core was at 100% with the PPU. Anyway the bonus maps are a waste of time. Its the main game that matters and the PPU is usefull for the main game.





”More like Quad Core - PhysX entirely = over 60 FPS.”
Impossible. If the dual is getting 12.1 fps then a quad is not going be 60.
quad core + PPU is always going be faster then just quad core.



“Uhm... why would you buy a PhysX card to not play the PhysX maps?”
For extra speed. Kind of the same reason you get a SLI setup or faster CPU. Buying a PPU just for PPU maps is normally silly. The maps are a waste of time, apart from to play around with once or twice



”sorry pottsey, but it hasn't change an iota. firstly, claiming 30fps when its not, 26fps would need close to a 20% speed boost to be counted as 30fps. 20% is not a small difference.”
It is with quad core on PPU maps and is on normal maps its over double 30.The game is CPU limited and quad core make a large diffrance.

A quad core can give more then a 20% FPS boost.





“the WHOLE point you keep making is that PPU negates the need for cpu power, so why would an ultra fast dual core make a difference.”
It only negates the need for cpu power for physics. If your CPU limited due to none physics a PPU helps but you will still be CPU limited. A PPU means you can get away with a weaker CPU on the normal maps. Not on the bonus maps.




“the whole point you miss here, is with a quad core, the ONLY difference will be on the PPU specific maps ppu speed will stay the same,!”
No it won’t. It will go up over 30%.
 
Last edited:
“Normal map without either = 60
Hmm yes I know what I want!!! 60 FPS Vs 30FPS with some slightly better physics hmm tough!”

Normal maps with PPU over 60. I know which I will take the PPU much better then without.

There are no disadvantages to having the PPU only advantages from the speed boost.




”Funny how Pottsey says the low framerates in Crysis are due to physics and not the GPU, yet the lower dramerates in the PPU maps are coz of the GPU/CPU. Please make your mind up Pottsey.”
Learn to read, I am being constant. I said the GPU/CPU is the reason for low FPS in Crysis. I said Crysis is both CPU limited and GPU limited in certain situations. During normal gameplay its mostly only GPU limited. I never said the GPU doesn’t factor into low FPS in Crysis dueing normal gameplay.
 
Crysis is both CPU limited and GPU limited in certain situations. During normal gameplay its mostly only GPU limited. I never said the GPU doesn’t factor into low FPS in Crysis dueing normal gameplay.
Will you please stop saying that Crysis is CPU limited? I've already proven to you that Crysis is only CPU limited at super low resolutions like 800x600 with lots of physics going on. In fact you can't even say that's CPU limitation as the graph goes up and down like a whore's drawers rather than remaining constant.

72j4owx.png


Crysis Benchmark: 800x600, all options low but physics high, no antialiasing.



Also: How you can call a Core 2 Duo and 8800GT mid-range is beyond me, it really is.
 
Pottsey said:
”More like Quad Core - PhysX entirely = over 60 FPS.”
Impossible. If the dual is getting 12.1 fps then a quad is not going be 60.
quad core + phsyics is always going be faster then just quad core.
I think you misunderstand, I meant Quad Core MINUS the whole Ageia PhysX junk results in 60+ FPS. What I'm trying to say is, why not just you know, not use PhysX since it seems to just hamper performance on the PPU maps? Get rid of the PPU only stuff entirely and you have a nice steady framerate. I'm sure the PPU does take some load off the CPU, but it still performs like ass either way so why not just junk the entire thing and stick with good old CPU based physics such as Havok and avoid this whole problem?

Pottsey said:
“Uhm... why would you buy a PhysX card to not play the PhysX maps?”For extra speed. Kind of the same reason you get a SLI setup or faster CPU. Buying a PPU just for PPU maps is normally silly. The maps are a waste of time not apart from to play around with once or twice
What exta speed? Where is your proof? You have none, and until you do I won't believe a PPU helps at all in non-PPU maps.
 
“Exqueze me? Not high end hardware?“
Quad core = high end
Dual core = mid end.

Unless I got the GPU’s muddled up its
8800 GT = mid end
GTS/GTX = high end.
Jesus you really dont know what you are talking about...

Normal map without either = 60

Hmm yes I know what I want!!! 60 FPS Vs 30FPS with some slightly better physics hmm tough!
And about £150 saving to boot...

“Uhm... why would you buy a PhysX card to not play the PhysX maps?”
For extra speed. Kind of the same reason you get a SLI setup or faster CPU. Buying a PPU just for PPU maps is normally silly. The maps are a waste of time, apart from to play around with once or twice
Icing on the rancid cake...

ps3ud0 :cool:
 
Last edited:
This is just another huge nail in the coffin that contains the PhysX tech. The thread title is misleading too - 100% performance increase in a PPU map with a PPU over a PPU map without a PPU, sure, it's technically correct, but both scenarios perform awfully, so the clear solution is to junk the technology and move on. It's like if a PPU map performed at 5FPS without a PPU and 10 with, that's a 100% increase but it's still completely unplayable.

I played the Warmonger game yesterday (Without a PPU, of course) to see what the big deal was. It looks awful and performs even worse than it looks. It's not worth buying a PPU for.
 
“Will you please stop saying that Crysis is CPU limited? I've already proven to you that Crysis is only CPU limited at super low resolutions like 800x600 with lots of physics going on.”
You haven’t proven anything as that CPU chart is without the situation I said that makes the game CPU limited. I already said during normal gameplay without extra physics the game is not CPU limited.

Its only CPU limited with the extra physics turned on with the ini file and lots of physics going off at once. The rest of the time it’s not CPU limited. Your chart is without those extra physics so your chart shouldn’t be CPU limited, hence your chart agrees with what I said. It doesnt prove me wrong.





“I think you misunderstand, I meant Quad Core MINUS the whole Ageia PhysX junk results in 60+ FPS.”
You quoted my scores of the PPU map and dual core then said quad core is 60fps. I thought you meant quad core on the PPU maps.




"Get rid of the PPU only stuff entirely and you have a nice steady framerate.”
Then add a PPU and you have an even better frame rate.




“What exta speed? Where is your proof? You have none, and until you do I won't believe a PPU helps at all in non-PPU maps.“
Apart from all the links I gave from end users in the others thread. I wish some website would do decent benchmark with and without the PPU using the same charts that where in the link I posted in this thread.
 
“Sorry Pottsey but how can you justify a 50% framerate drop in a PPU map over a non PPU map an advantage?”
As you get far more then a 50% drop without the PPU so its an advantage. Anyway as I said the PPU maps are rubbish. There are tons of things wrong with them. It’s the normal game that matters and the PPU is an advantage for that.
 
You haven’t proven anything as that CPU chart is without the situation I said that makes the game CPU limited. I already said during normal gameplay without extra physics the game is not CPU limited.
If a game only utilises 100% of the CPU in exactly one situation, one I might have to add that isn't even in the game but has to be heavily tweaked in the editor, then it's not CPU limited.

Your chart is without those extra physics so your chart shouldn’t be CPU limited, hence your chart agrees with what I said. It doesnt prove me wrong.
No, what my chart shows is that even when utilising heavy amounts of physics that are actually in the game, even at 800x600 resolution, you don't get 100% utilisation of the CPU.

What this means, I'm sorry to break it to you, is that the CPU doesn't break a sweat with lots of physics. It might not be up to scratch when the physics that the Ageia API produces are in effect, but as far as Crysis is concerned there's no problem here.

P.S. If your benchmarks are anything to go by, it seems that the Ageia PPU isn't up to the task either when it comes to physics produced by the PhysX API. Maybe the people in charge of level design need to go back to the drawing board, because it's either that or the Ageia PPU sucks.
 
Can someone please expain how taking the normal physics from a quad core and move to a PPU increases the fps? I mean, there are hardly "cpu intensive" are they.
It's because the CPU already has plenty to do. And because Ageia PhysX affects are so pointlessly heavy on resources that, as we can see from the UT3 benchmarks, even the Ageia PPU can't keep up.
 
Pottesy, you really arn't winning.

IF it was good, people would buy it. It isn't so people don't!

We're not saying its **** just cos we feel like it, were saying it COS IT IS!
 
Back
Top Bottom