PhysX FluidMark results thread

[Tatty_Two]

tatty what CPU are you running also what speed was this bench ran on? need it for score list

Q9650 @ just 3.825gig for the bench, she will do 4.5gig flat out but am saving a fluidmark run at that speed until someone beats my current score!

 

[p_o_s_pc]

thanks. I updated the list

 

[lemonadesoda]

The results of the REAL Asus P1 PhysX Card sux!!

 

[p_o_s_pc]

+1 they do but before nvidia started this that is all that you could get

 

[lemonadesoda]

Why do you insist on GPU-z,CPU-z when the key info is already shown in the Fluidmark screen?

 

[p_o_s_pc]

Why do you insist on GPU-z,CPU-z when the key info is already shown in the Fluidmark screen?

because fluidmark doesn't always get it right and it is just nice to see the same thing from more then one thing

 

[soldier242]

tatty your scores are amazing, i should put my gtx on water to achive some higher clocks ... but
i think even with clocks like yours i won't beat ya ... strange, must be xp64 or the low cpu clock or everything

 

[Tatty_Two]

tatty your scores are amazing, i should put my gtx on water to achive some higher clocks ... but
i think even with clocks like yours i won't beat ya ... strange, must be xp64 or the low cpu clock or everything

It's all in shader speed to be honest coupled with cpu speed of course, I am saving a 4.4gig CPU overclock run for when someone beats me!, try lowering your core clock down but whacking your shaders up, you should see a decent difference, Px is predominantly (but not exclusively) shaders, ohhhhh and of course, if you have a 64bit operating system, forget it, it a 32bit bench, you will still get a score of course but nowhere near what it could be.

 

[GrandAdmiralThrawn]

nVidia PhysX

I just registered here, because I was looking for information regarding my FluidMark Score. First, my System Config:

HW:
==
Intel D975XBX2 Rev.305
Intel Core 2 Duo E8500 @ Stock Speed (3166MHz)
4GB Corsair DDR-II/1000 5-5-5
Creative SoundBlaster X-Fi XtremeMusic
3ware 9650SE-8LPML RAID-6 Controller
BFG GeForce 280 GTX @ 632/1296/2322

SW:
==
Windows XP SP3, fully updated
ForceWare 178.13 + PhysX 8.09.04 (the new WHQL package, drivers fully uninstalled beforehand with assistance of DetonatorRip and NFR)
No FSAA, no AF for this test, VSync forced OFF
FluidMark 1.0.0, 640x480, 60secs


The system's pretty "fresh", this is only the second ForceWare installation here.

First, I noticed that ANY CPU hog running alongside of FluidMark will kill the score, so i completely freed up my CPU, no more SETI@Home, no more AV running..

Now, the result is around 700 in Software and around 3600 using Hardware PhysX. I just don't get it.. 3DM06 Score sits slightly above ~15.000.

Now, question is, what am I doing wrong here.. I also noticed, that when running FurMark for instance, the loud fan of my 280 GTX starts to spin up like hell almost immediately. However, when running FluidMark, the Fan stays silent and the card below 80°C...

Oh, I also tested SuperPi and Everest too, SuperPi Mod 1.5 gave me ~14 secs for 1M and Everest gave me 7GB/s mem writes, so i guess those components should not be the cause?

Am I forgetting something?

Help appreciated, thanks a lot..

 

[Tatty_Two]

I just registered here, because I was looking for information regarding my FluidMark Score. First, my System Config:

HW:
==
Intel D975XBX2 Rev.305
Intel Core 2 Duo E8500 @ Stock Speed (3166MHz)
4GB Corsair DDR-II/1000 5-5-5
Creative SoundBlaster X-Fi XtremeMusic
3ware 9650SE-8LPML RAID-6 Controller
BFG GeForce 280 GTX @ 632/1296/2322

SW:
==
Windows XP SP3, fully updated
ForceWare 178.13 + PhysX 8.09.04 (the new WHQL package, drivers fully uninstalled beforehand with assistance of DetonatorRip and NFR)
No FSAA, no AF for this test, VSync forced OFF
FluidMark 1.0.0, 640x480, 60secs


The system's pretty "fresh", this is only the second ForceWare installation here.

First, I noticed that ANY CPU hog running alongside of FluidMark will kill the score, so i completely freed up my CPU, no more SETI@Home, no more AV running..

Now, the result is around 700 in Software and around 3600 using Hardware PhysX. I just don't get it.. 3DM06 Score sits slightly above ~15.000.

Now, question is, what am I doing wrong here.. I also noticed, that when running FurMark for instance, the loud fan of my 280 GTX starts to spin up like hell almost immediately. However, when running FluidMark, the Fan stays silent and the card below 80°C...

Oh, I also tested SuperPi and Everest too, SuperPi Mod 1.5 gave me ~14 secs for 1M and Everest gave me 7GB/s mem writes, so i guess those components should not be the cause?

Am I forgetting something?

Help appreciated, thanks a lot..

Well in a nutshell, your shaders are running slow, but that cant be the only answer, your still slow and should be quicker....Px is basically about shader power, I hope that is 32bit Windows XP?

 

[GrandAdmiralThrawn]

Windows XP

Normal 32-Bit XP, yes.

Shaders are Stock, but still.. Every 8800 beats me.

By the way: Is the 3DM06 score of ~15.000 ok? I didn't run 3DMark for a pretty long time, last time I was comparing results was back with 3DM01... Just trying to find out if this is somehow a graphics card issue or a PhysX issue..

Edit: I'm suspecting the card doesn't enter "Performance 3D" mode here, and keeps running with 2D or Low-Power 3D clocks. May that be a cause? The fan not spinning up keeps me thinking..

 

[Bytor]

Ahahahahahaa I'm dead last.....And the only ATI guy on the list.....

Oh well its a Nvidia bench anyway... Nice scores guys...

 

[Tatty_Two]

Normal 32-Bit XP, yes.

Shaders are Stock, but still.. Every 8800 beats me.

By the way: Is the 3DM06 score of ~15.000 ok? I didn't run 3DMark for a pretty long time, last time I was comparing results was back with 3DM01... Just trying to find out if this is somehow a graphics card issue or a PhysX issue..

Edit: I'm suspecting the card doesn't enter "Performance 3D" mode here, and keeps running with 2D or Low-Power 3D clocks. May that be a cause? The fan not spinning up keeps me thinking..

Every 8800 beats you because their stock shader speed is higher, of course as I said thats not the complete answer, something else is wrong here, let me give it some thought.........

 

[Tatty_Two]

Ahahahahahaa I'm dead last.....And the only ATI guy on the list.....

Oh well its a Nvidia bench anyway... Nice scores guys...

Yes, so you will probably be running in software mode.

 

[Mussels]

Thrawn: 15,000 in 06 is slow as well.

You wouldnt have Vsync, antialiasing or anisotropic filtering forced on in the forceware panel would you?

 

[GrandAdmiralThrawn]

Scores

Unfortunately, no. VSync is forced off in nVidia CP, and so are FSAA and AF.

8800 shader clock is higher, but shouldn't my 240 shader processors make up for this? Or does PhysX only run on a limited amout of shader units?

Well, whatever, if my 3DM06 result is also too low, something else must be wrong, but I can't figure out what...

 

[Tatty_Two]

Unfortunately, no. VSync is forced off in nVidia CP, and so are FSAA and AF.

8800 shader clock is higher, but shouldn't my 240 shader processors make up for this? Or does PhysX only run on a limited amout of shader units?

Well, whatever, if my 3DM06 result is also too low, something else must be wrong, but I can't figure out what...

In that case, your GPU is bottlenecked by your CPU, you have a powerful card and I would guess you would not get the full performance out of it on anything less than about 3.6gig........anyways, games is where it really matters, remember these test are "synthetic" and dont necessarily give a real world gaming indicator.

For example, with my quad at 4.4gig and a heavily overclocked GTX260, I get over 22,000 in 2006, that overclock not only improves the CPU score, it really bangs up the SM2 and SM3 scores as well.